SuDiBe Tandemseiten 


Disc Brakes

by Bill McCready

(Scheibenbremsen für Tandems - sehr viel Hintergrundinfos und ein wenig Werbung. Aufbereitet aus der englischsprachigen Mailingliste, 11/2003)

This LONG post contains the following revelations. If you read it all, you'll discover the following and a lot more.

Santana has tested brakes for years. Our protocol has been published (on T@H) and distributed to various disc brake manufacturers (including Hope, Formula, Shimano, Avid, Magura and Hayes).

Our short, steep test hill (Mountain Ave. in Claremont, CA) is a wide, smooth roadway. From the dead end at the top to the bridge at the base is a distance of .67 miles (1,074 meters). The drop in elevation is 528 feet (161 meters). While the steepest portions exceed 20%, the entire grade averages 15%. Our "test course" is open to everyone (schoolkids ride it daily). Besides the hill itself we use 400 pounds of riders, a test brake mounted to the rear wheel (which can be rim, disc or drum), a front standby brake, plus calibrated "blister" tapes and a digital infrared "gun" to measure heat. We use the test brake to coast down the grade at 15 miles per hour. At two intermediate locations, a full application of the test brake is used to bring the tandem to a full stop. The test brake is immediately released and, after a 5-second pause, the bike is allowed to coast back up to 15mph. The test brake is then applied to maintain this speed. If the brake is still functioning at the bottom of the hill, it is fully applied for a third time. As soon as the bike stops, we take heat measurements. We've now tested nearly two dozen disc brakes. We have always invited manufacturers to observe our tests (and Hope once sent a representative). If any of you would like to attend a future session, send us an e-mail.

Because this is a pass/fail test of a brake's heat capacity, stopping distances are not recorded. The entire test is completed within ten minutes. In order to pass, the brake must (1) maintain the 15mph test speed between stops, (2) bring the bike to three complete stops, (3) not stick closed, and (4) not need new parts. Sound easy?

The Arai drum brake passes, as does a tandem-specific version of the Formula hydraulic disc (with either 185mm or 203mm rotors). No other brake has yet passed. How do they fail? Cheap hub brakes (drum and band) and some discs will fail before the first interim stop. Rim brakes and most disc brakes will fail at or before the second stop. (Rim brakes don't actually fail, but the tire leaves the rim). Shortly after the second stop the 8-inch mechanical discs from Avid and Hayes have lost all power (through fade), and can no longer control downhill speed. After the standby brake is used to rein in the resulting runaways, we note melted rubber seals (Hayes) and plastic adjustment pieces (Avid). In other words, these highly rated "downhill" brakes with 8-inch rotors fail two ways: they can't control downhill speed AND they melt in the process.

When I heard that Avid had revised their longstanding "not for tandem" rating, I contacted them and spoke with Paul Kantor.

Q: Have you made any improvements?
A: No.

Q: Have you done any testing?
A: No.

Q; Why the change?
A: Cannondale and Todd Shusterman (Da Vinci Designs) have told us that our brake has passed their tests.

Q: What were those tests?
A: We don't know.

PS: Before someone accuses me of a hatchet job, they should realize that Santana has, for many years, installed Avid rim brakes on nearly every tandem we deliver. Because most tandem buyers (and bicycle mechanics) are wary of hydraulics, Santana would love to offer a capable mechanical disc, and will work with Avid (and others) to develop a non-hydraulic disc that is good enough for tandems.

Part 2: A History of Disc Brake Failures

Because tandem brakes have overheated with disastrous consequences, Santana has tested a brake's heat capacity for over 20 years. During that time a number of manufacturers have produced unworthy brakes that were eagerly purchased by tandem enthusiasts.

Most memorable to me was a decade-long disagreement that existed between Santana and Phil Wood over the performance of his 1975-1984 mechanical disc brake. While the design of his lightweight brake was inspired---and Phil actively promoted its use for tandems---we quickly discovered that the brake was failure prone. Phil not only ignored our complaints, he flatly denied that a "real" problem existed. In the winter of 1980, when John Schubert and Gary Fisher arrived in Southern California to test six tandems for Bicycling Magazine, they wondered why Santana didn't use the Phil disc found on other brands. At the time a Phil rear disc was not only the most popular option on custom tandems, Schwinn had removed rim brakes to equip their Paramount tandems with dual Phils instead. The following day these two editors (who didn't want to believe my story) failed a brand new Phil disc.

Was the failure reported? No. After Phil convinced someone at Bicycling that the failure was a fluke, that particular test result was edited out.

Later, after at least three tandem teams experienced simultaneous dual disc failures, and Phil settled at least two lawsuits (one, unfortunately, where a stoker was seriously and permanently disabled), he quit selling the brake or any of its spare parts. Within a couple of years a thousand (?) of these expensive brakes were retired when the owners could no longer buy replacement pads. In today's era of CPSC-mandated recalls, Phil Wood would have needed to spend a quarter million dollars to buy-back these brakes. (And Schwinn would have needed to retrofit two-years of Paramount tandems.)

Is the Phil incident unique?

No. Over the years Santana has rejected nearly two dozen disc brakes. A more recent example was the 1990-96 Hope mechanical discs. While these brakes were merely ineffective, custom tandem builders installed hundreds without question. Today, in spite of gushing reviews and glowing testimonials from Hope's early customers, almost all of these Hope mechanical discs have quietly been removed.

Which brings me back to Avid.

Last summer, when Avid reportedly removed their prohibition on installing their discs on a tandem, I called them and was surprised to learn that their change of heart was not based on product improvement or internal testing. Instead, their new direction was the result of reports they had received from Cannondale and Da Vinci.

P.S. By now you realize that this post is too long for a quick read. Instead of skimming this multi-part post, why not print it (using recycled paper!) or save an electronic copy.

PPS: If, on the other hand, "A post from Bill..." causes boredom or anger, use the delete key!

Part 3: In search of valid test

When I asked Avid's Paul Kantor why Avid was no longer warning customers that their disc was unsuited for tandems, he replied that test reports from Todd Shusterman (Da Vinci Designs) and Cannondale had caused Avid to abandon their long-standing position.

When I phoned Todd Shusterman in September to talk about disc brakes, he was surprised to learn that Avid had credited him with a test. Actually, he had reported to Avid that when and his wife had descended France's Mount Ventoux using a rear Avid disc and front rim brake. His report to me included the following:

While descending Ventoux's west road Todd had "let the bike run" on straightaways. He stopped part way when he needed to readjust the pads on his rear disc. Upon reaching the town at the base of Ventoux, he found he'd used his front brake enough to get his rim "too hot to hold." He didn't have a touring pack, and says his team weighed 320 pounds.

My comments: Ventoux's west road is not as steep or twisty as the classic southern approach used in the Tour de France. Although the steepest sections of this newer and wider road exceed 9% (according to Michelin), the average gradient is a less remarkable 7.5%. The long straightaways on this route allowed Todd's brakes to cool between turns. A mid-way stop (to adjust the disc pads) allowed additional cooling. His use of the front rim brake further protected his disc. While the Shustermans should be admired for their day's ride (a tandem ascent of Ventoux from any angle is an impressive feat) their westward two-brake descent at moderate speeds was not a severe test of a tandem's disc brake.

While Todd was pleased with the performance of his disc, he did not imagine Avid would interpret his report as a "real test."

At Cannondale I reached Mark La Plante, a 20-year veteran who runs their testing facility. The two of us have now shared four long phone conversations about disc brakes and tandems. A couple of years ago Cannondale devised a tandem-specific lab test for a disc brake that would simulate a descent of New Hampshire's Mount Washington. Their calculated "braking torque" to descend the 7.6 mile, 12% grade at a safe constant speed was 60 foot pounds. While Cannondale hasn't used this test for rim or drum brakes, they've tested a number of discs. Amazing (to me) was that Cannondale and Santana tests contained surprisingly similar elements:

Both tests rely on one brake. Gross test weight: Santana 435#, Cannondale 400#. Test speed: Santana 15mph, Cannondale 15.5mph. Ambient test temp: 70 degrees F (plus or minus 10F). Cooling: Santana outside air, Cannondale electric fan.

The biggest difference in the test: Mount Washington: 12% for 7.6 miles. Mountain Avenue: 15% for 2/3rds of a mile.

Q: How many discs have passed Cannondale's "Mt. Washington" standard?
A: Zero.

Q: What about the Avid?
A: Avid's disc failed seven minutes into the 30-minute test.

PS: A T@Her who attended Interbike and stopped by the Avid booth, reported that he had talked to an "Engineer." While Avid has a capable design staff headed by their President, Wayne Lumpkin, they've never had an engineer, test facility or testing program. Their ball-bearing disc is produced by Wellgo, a respected Taiwanese company famous for its pedals. Few Taiwan component makers have test facilities either, as a nationally subsidized test center is a function of the Taiwan Trade Commission. While the Trade Commission's test center is quite good, as far as I know they have not as yet developed specific test criteria for disc brake performance. Santana and Avid have discussed some potential remedies for the failure points we have found, and our companies will work together this winter to increase the heat capacity of Avid's current design (which is limited due to teensy brake pad dimensions and inadequate brake pad airflow---factors that are 3-4 times less critical on a design that was, after all, intended solely for single bikes).

PPS: Following years of frustration with Asian suppliers, Avid turned to Italy and entrusted their HYDRAULIC disc brake program to Formula, an experienced producer of motorcycle disc brakes. Formula has engineers, test facilities, a race-support program and lots of hard-earned knowledge. Because of Formula's development input and manufacturing expertise, Avid's just-introduced "Juicy" hydraulic looks very promising (for single bikes).

Part 4: Cannondale's disc brake testing

Two years ago Cannondale's Testing Engineer, Mark La Plante, designed a tandem specific lab test to simulate a descent of New Hampshire's Mount Washington (home of an annual bicycle challenge). After every disc Cannondale tested failed this test, they cut their tandem brake standard in half. Instead of applying the lever to achieve 60 foot-pounds of braking force, with their new test the brake lever would be squeezed half as hard in order to create 30 foot-pounds of braking force. Still 30 minutes long, they call their new test "Half Mt. Washington." (To put the 30 and 60 foot-pound figures in perspective, a strong squeeze of a brake lever can create more than 120 foot-pounds of braking force).

When I asked Mark La Plante why Cannondale had decided to make their test twice as easy, his reply was that without this change Magura's "tandem-approved" Gustav-M could not pass. While the original "Mt. Washington" test may have been a reasonable real world standard, Cannondale was reluctant to fail a "certified" brake from a respected manufacturer. By cutting their test in half, the Gustav-M could pass. Once the bar was lowered, Magura's Julie also passed. And if Cannondale relaxed their standard a bit further by allowing the test to be paused for pad readjustment, Avid's largest 8" (203mm) disc could also pass.

I then contacted Buck Mitchell from Magura USA to purchase a 3rd-generation Gustav-M, and to ask about their "tandem approved" rating. When the answer arrived from Germany, I learned that Magura tests brakes for both power and heat. To pass their tandem-specific heat test a brake needs to produce 1000 watts of stopping power for 15 continuous minutes without boiling fluid or suffering from excessive fade. And, after cooling, it needs to repeat this performance without needing readjustment or using up more than 30% of its pads.

Here is a quick summary of tandem-specific disc testing:

Santana: 15% grade for 2/3rds of a mile.
Cannondale: 30 foot-pounds of braking force for 30 minutes.
Magura: 1000 watts for two tests of 15 minutes each.
Hayes: Subscribes to Santana's test.
Hope and Shimano: No tandem-specific test standard.
Avid: No testing whatsoever.

PS: It is interesting to me that Cannondale's reason for cutting their original tandem test standard by 50% was to let a respected Magura disc pass. But then, in order to also "pass" an Avid disc, Cannondale watered down its 50% standard to allow a mid-test readjustment---a "fudge" that is not allowed by Magura's test standards.

PPS: Watered-down testing reminds me of two years spent on a local school board. If you can't get students to pass, the tests are too hard!

Part 5: Avid denies "tandem certified" and Cannondale works on a new tandem disc test.

After Avid reportedly changed their long-standing prohibition on installing their disc brakes on tandems, I gave them a call. Instead of a design change or test data, however, Avid referred us to Da Vinci and Cannondale for their test results. Instead of a test, Da Vinci's Todd Shusterman related a successful tandem descent of Mount Ventoux (13 miles averaging 7.5%). Cannondale's initial test was a simulation of Mount Washington (7.6 miles averaging 12%). When the Avid brake failed this test, Cannondale cut the required braking force from 60 foot pounds to 30 foot pounds. When the Avid disc failed the "Half Mount Washington" test, Cannondale amended the protocol a second time to allow a pause for readjustment. Based on passing this third version of a tandem-specific test, Cannondale's Product Managers were allowed publish a tandem specification that included Avid discs. In the meantime, Cannondale's Testing Department was unaware of Avid's still-in-place ban on tandem disc installation.

Because Santana had not yet tested the newest 8-inch (203mm) version of Avid's "not for tandems" brake, in September we scheduled a retest to see if Avid's biggest brake would survive Mountain Avenue (2/3 of a mile averaging 15%). In case you've forgotten, three brakes have passed our six-year-old test: a standard Arai drum and two tandem-specific discs from Formula (185mm and 203mm).

When tested, Avid's 8" disc failed after one-half mile. As is sometimes the case, this test was a double failure---the brake could not maintain the 15mph-test speed (excessive fade) and was disabled in the process (melted adjustment fittings).

Instead of moving the finish line (!), we decided to mate a second Avid caliper (the first one could no longer be adjusted) with a prototype 10" rotor. Unfortunately, at the half-mile point the Avid again experienced "runaway" fade. Then, before the tandem could be brought to a stop, our prototype 10-inch rotor warped. (We later learned that our prototype, a "show-only dimensional sample," hadn't been heat-treated). Because the adjustment pieces again melted, this test was a rare triple failure---warped non-Avid disc, excessive fade and melted pieces.

When I visited Avid's Interbike booth two weeks later, Paul Kantor was not surprised by the failures. He then "reminded me" that Avid has never "certified" the brake for tandem use!

Surprised by his remark, I asked for clarification. Kantor replied that with tandem builders (such as Da Vinci and Cannondale) ignoring Avid's ban and performing their own tests, Avid has decided to abandon their former position, and will let builders, dealers and customers make their own decision. Avid's only change, Kantor continued, was to take "no official position" on tandem use. Kantor then directed me their current OEM product sheet and expo signage. While recommended uses for Avid's mechanical discs include cyclo-cross, touring, and recumbents, the word TANDEM does not appear on their list of certified applications.

What about using the disc as a drag brake? Avid's installation instructions, while silent about tandem use, specify that the cable must be attached to a "brake lever." Kantor, not convinced that anyone should install an Avid disc on their tandem, thinks it doubly inadvisable that someone would attach it to a shift lever.

When I spoke with him last week, Cannondale's Mark La Plante was in the midst of developing a new tandem-specific disc brake protocol. After reviewing the issues, La Plante believes that the 30 foot-pound braking force that allowed the Avid brake to pass is too low. He has decided to use a real tandem and real hill (instead of a theoretical simulation) to develop test #4. As a starting point he sent a 400-pound test team and instrumented tandem to Blankly Hill to obtain some real world brake force ratings. In the meantime, the Avid spec for Cannondale tandems is on hold. While Cannondale's Product Team would undoubtedly like to see the Avid brake re-qualified, Mark fully realizes that because Avid has not approved their disc for tandems, Cannondale will be forced to accept a higher degree of responsibility for any liability claims and/or CPSC recalls that could occur as a result of incidents where inadequate braking is alleged.

P.S. As the concluding chapters are not less technical, you might consider printing or archiving this multi-part post so that you can re-read it in its entirety. And in case you're wondering, because all of these installments were completed before the first one was posted, a planned follow up post will (in a week or so) address the unanswered questions that are bound to be raised.

Part 6: Is Santana's test too hard?

For years Avid had advised builders, dealers and customers not to install their disc brake on a tandem. When we tested the brake, we agreed. Our tests confirmed rapid fade to ineffectiveness, melted components and the occasional warped rotor.

Yet here on T@H, even before Avid relaxed its position, you could find the following kinds of posts (paraphrased from memory).

Paraphrased post #1: "I've been using a rear Avid on my recumbent single and am totally satisfied. I don't care what they say, if I owned a tandem I'd install an Avid on it."

My comment: Some people skipped Physics. Tandems are, on average, twice as heavy. Because of a superior weight-to-resistance ratio, on steep hills a tandem can coast 40% faster than a single (and probably 20% faster than the typical recumbent). Because a twice-as-heavy bike travelling 40% faster has four times more energy (essentially the same as "Energy equals mass times velocity squared") a tandem brake will need to create (and tolerate) up to four times more heat than a well-designed brake for a single bike. Imagining that a brake will work on a tandem because it doesn't fail on a recumbent is imprudent.

Paraphrased post #2: "Instead of Santana's contrived test, our company performed a real-world test. We gave an Avid-equipped tandem to a local team that wins races. They rode it over a big mountain. They told us the Avid performed great."

My comment: From the "wins races" description we can imagine that the testers were (1) lightweight and (2) fearless---qualities that won't generate downhill heat. As for the test course, we don't know if the descent was particularly steep---or even twisty enough to require constant braking. Although this builder's post included a phone number, he did NOT add, "Operators are standing by to accept your order." Because buyers of expensive tandems want disc brakes, every tandem builder (including Santana and Cannondale) has a built-in incentive to avoid meaningful tests. Is Santana too prudent? Or are other builders too eager to take your money?

Paraphrased post #3: "Our 310 pound team regularly descends a local hill and uses the Avid to scrub off over 30mph of excess speed before diving through a corner. After maybe 20 repeats, we haven't even needed to replace the pads."

My comment: A repeated test at a moderate level should not fail any brake. In fact, the original Mafac rear brake on my first tandem would easily pass this test. While numerous decelerations from 50 to 15mph might seem impressive, a demonstrably tougher test would be a single emergency stop from 50mph. I know from controlled test results that an Avid disc brake will not be able to stop a 400-pound team from 60mph. How do I know this? In a bench test performed at Cannondale, a strong one-hand brake application through a standard lever and cable caused an Avid brake to fail from excessive heat within five seconds. In Santana's test, where the speed never exceeds 15mph, the Avid performs much better (and survives a half mile of 15% descent with two stops before failing).

Paraphrased post #4: "My Avid brake performs great. With a 100 pound stoker we can skid the rear tire."

My comment: The reason for putting a disc at the BACK of the tandem has little to do with braking power. Further, when a powerful front brake is applied simultaneously, a rear skid (even with a heavy stoker) can be a misleading indicator of disc brake performance.

Confused? It's easy to confuse stopping power with performance. While a stick through the spokes, for instance, creates awesome stopping power, the modulation leaves something to be desired. A heavy tandem team that will never leave Florida may need a more powerful brake (to avoid inattentive drivers turning left), than a lighter team that lives in Switzerland (where drivers pay attention). While the Swiss couple should also want stopping power, a more pressing need for tandeming through the Alps is heat dissipation, a quality not needed in Florida. Ignoring extreme cases, the best answer for the average road tandem team is a powerful brake for emergency stops, and a heat-resistant brake for steep descents. While the powerful brake MUST be in front (where 80% of your emergency stopping power can be generated), the heat dissipating brake is equally effective at either end. And though many tandem riders have come to love their "drag brake," if your heat-dissipating brake also has adequate efficiency (hydraulics and larger rotors both help) you'll need only two brakes to descend the longest and steepest hills without a hand cramp. If you've followed the logic thus far, you may now realize that the most efficient tandem set-up will be a powerful rim brake up front and a heat-tolerant disc brake in back.

Final Installment: Why not put dual discs on a road tandem?

While those who want a flashy tandem will continue to mount discs to the front of a road tandem (I should know, my original tandem has sported dual discs since 1975), the resulting tandem will be less efficient in 8 ways. Before listing the problems, let's review the physics.

Most bicycle enthusiasts approach this issue based on their knowledge of cars and motorcycles. If the best cars and motorcycles have all switched to discs, they wonder, aren't discs an inevitable step forward for bikes?

Maybe not. Over the past 30-odd years cars and motorcycles have evolved from (smaller) drum brakes to (larger) disc brakes. The best bicycles, however, have always relied on rim brakes. And, when you think about it, a bicycle's rim brake is a disc brake. That's right; those normal-looking sidepulls, center-pulls and V-brakes have always been disc brakes. More important, the rotors (or rims) are relatively large and, therefore, extremely efficient. As crude as they seem, on a pound-for-pound basis your bicycle's rim brake system is more powerful than the disc brake system on your car.

But on a bicycle, isn't a disc brake more powerful than a rim brake? No. With the same amount of hand pressure the rim brake provides more foot-pounds of braking force.

How can that be? Because stopping power is a squared function of braking radius, an 8-inch rotor (with a braking radius of just under 4 inches) is a poor match for a 700c front rim (with a braking radius of over 12 inches). Here's the math: Since 91mm squared equals 8,281, and 311mm squared is 96,721, a caliper pinching a 700c rim can produce nearly a dozen times more stopping force than an identical device pinching an eight inch rotor. While bicycle disc brakes have come a long way (and are improving at a rapid pace), today's disc calipers have not yet overcome this 12x disparity.

On bikes with 26-inch wheels, however, the rim-to-rotor advantage drops to 9:1. This, plus the difficulty of mating rim brakes with suspended wheels, helps to justify the installation of front discs on mountain bikes, where mud is a major consideration. Disc brakes on road racing singles? Don't hold your breath!

Because disc brakes LOOK powerful and high-tech, many enthusiasts dream of a tandem with dual discs. While I whole-heartedly agree that a capable REAR disc is an immense step forward (at least for moderately heavy teams that will venture into the mountains) a road tandem with a front disc brake will suffer 8 inefficiencies:

First, unless the team is very light, today's discs aren't powerful enough for the front end of the tandem (where 80% of the tandem's emergency braking power must be generated). Second, the bike will need a wider fork or a weaker dished front wheel. Third, in either case the fork will need to be considerably heavier (80% of your tandem's stopping power will be applied to the base of one fork leg). Fourth, the heavier (probably steel) fork will be less comfortable than today's carbon forks (carbon can't tolerate the heat of a tandem's disc brake caliper). Fifth, the wheel (dished or not) will need a heavier hub. Sixth, the disc brake itself is also heavier. Seventh, if you want a sweet-handling bike, hanging an extra pound of weight on one side of your steering axis won't help (an asymmetric application of braking torque won't help either.) Eighth, the dollar cost of the previous inefficiencies is over $300.

If, in spite of the above listed reasons, you still think a dual disk road tandem is cool, remember that the only reason a road tandem NEEDS a disc brake is heat dissipation, a function that can be performed by a lone disc at either end of the bike.

PS: Our next round of tests will include the newest Gustav-M with 190mm rotor, the new version of the Hayes mechanical (with caliper cooling ports), and an Avid mechanical disc mated with a refined 255mm (10-inch) wave-rotor produced by Galfer. Send an email if you'd like to attend!

PPS My first twofer (bought used in 1965) was a short-coupled T. Parsons British road-racing tandem that still wears silk sew-up tires. The dual discs I later installed were way cool. A cable from one Campy brake lever controlled a finned aluminum master cylinder mounted to the frame via special braze-ons. A pair of clear hydraulic hoses (with bright PINK fluid) ran to either wheel, where massive Hearst-Airhart competition go-kart calipers (also finned) clinched stainless discs mounted to custom chromed hubs.

PPPS: And when I needed to stop in a hurry, the other Campy lever worked a front rim brake!

D. Bettge; letzte Änderung: 10.11.2003