From: pardo@cs.washington.edu Date: Thu, 05 Nov 92 11:14:13 -0800 Subject: Drum brakes (WAS: The Spirit of Human Effort) Rim brakes operate on the ground with 1:1 leverage. Hub brakes have leverage that changes with the wheel diameter. A very modest hub brake can be effective on a small-diameter wheel but inadequate for a large- diameter wheel. I prefer hub brakes because they are more predictable in poor weather and are immune to rim damage. On tandems and other heavy vehicles, they don't tend to cause blowouts the way rim brakes do. However, hub brakes tend to weigh more and I have more problems with brake fade when using hub brakes. That's a few considerations... ;-D oN ( The brake in period ) Pardo From: R.Stclair@EBay.Sun.COM (R D St.Clair) Date: Thu, 5 Nov 92 13:44:01 PST Subject: Re: Drum brakes (WAS: The Spirit of Human Effort) In addition to Don's many good points... A rim brake acts directly on the rim, not putting much additional stress on the spokes, while the force of a drum break acts through the spokes. This may require a more agressive cross spoking pattern though I don't claim to have any hard data on the subject. Thanks, R.D. From: praetzel@marconi.uwaterloo.ca Date: Thu, 5 Nov 92 17:31:06 -0500 Subject: Re: Drum brakes (WAS: The Spirit of Human Effort) >A rim brake acts directly on the rim, not putting much additional stress >on the spokes, while the force of a drum break acts through the spokes. >This may require a more agressive cross spoking pattern though I don't >claim to have any hard data on the subject. Don't worry about it. Jobst's book went thru this. Basically the compression forces are on the order of 100 lb / spoke (guess off the top of my head since I know that the total forces are on the order of a ton). The additional forces of pedal, hitting bumps and braking are insignificant. - Eric From: " pardo@cs.washington.edu" Date: Thu, 05 Nov 92 14:44:46 -0800 Subject: Re: Drum brakes (WAS: The Spirit of Human Effort) >[Hub brakes put more load on the spokes.] Radially-spoked wheels certainly cannot handle the load of a drum brake. My nomination for most shocking picture of the year is the dust cover of _Richards' Ultimate Bicycling Book_ (I tink that's the title), which shows a radially-spoked wheel with a hub brake. Generally, though, spokes fail from fatigue, not from overloading. For appeal to authority see Jobst Brandt's _The Bicycle Wheel_ and his numerous postings on `rec.bicycles'. Fatigue failures are a weakening of the spoke that come from constant *changes* in the loading. The dominating factor, in this case, is the weight of the rider, applied once per revolution of the wheel. A nonintuitive consequence is that spoking the wheel tighter can improve spoke life, because the load (relaxation, actually) is spread across more spokes and, thus, the peak tension change per revolution (the change on each spoke) goes down with a tightly-spoked wheel. What's that got to do with drum brakes? In normal circumstances, the hub brake does not apply forces that will break spokes. To be sure, it does change the spoke tension while it is applied, but the change in tension is modest and is spread across ~9 spokes. It would be ~18, but the hub shell doesn't generally transmit torque. By contrast, the once-per-revolution tension change of rolling down the road are spread across maybe 4 spokes and the change is repeated once per revolution. >From all the above you might conclude that spokes should fail periodically and approximately randomly. In point of fact, they break when peak loads are applied: the brake is on, the bike is going over bumps, the bike is loaded, etc. They break because they have exceeded their capacity to hold load. But what is important is that the spokes *used* to be able to hold that load and have gotten weaker from fatigue. And that's what defines the limits. ;-D on ( An outspoken effort to drum up support ) Pardo From: " pardo@cs.washington.edu" Date: Thu, 05 Nov 92 17:06:03 -0800 Subject: Re: Drum brakes (WAS: The Spirit of Human Effort) praetzel@marconi.uwaterloo.ca writes: >[100lb/spoke, negligible compared to braking is insignificant.] I believe that the forces that Jobst addresses are those from *rim* brakes. The forces from hub brakes are several times those of pedaling and shouldn't be brushed of quite so lightly. Suppose you're braking 100kg of rider at 1g with a 700mm diameter wheel and that the tangent of the spokes is 50mm diameter. Then the tangent force exerted at the flange is about 14 times the tangent force at the tire. That's about 1400Kg divided among ~9 spokes is 155kg compared with a resting tension of about 50kg. That's *quadruple* the typical spoke load, nothing to sneeze at. Fortunately, deceleration rarely exceeds 0.5g, so the added load is `only' 75Kg, and with a good stiff hub there's double the spokes so tension is `only' ~35-40Kg, but that's still approaching double the static load on the spokes. ;-D on ( Getting drummed out of the score ) Pardo From: "Nickolas E. Hein" Date: Thu, 5 Nov 92 14:14:29 PST Subject: Re: Drum brakes (WAS: The Spirit of Human Effort) > Inaddition to Don's many good points... > > A rim brake acts directly on the rim, not putting much additional stress > on the spokes, while the force of a drum break acts through the spokes. > This may require a more agressive cross spoking pattern though I don't > claim to have any hard data on the subject. > > Thanks, > Thanks for the information. I don't know if I mentioned it, but the front wheel that has the drum brake is 16" diameter so the cross-spoking is pretty aggressive to begin with. Ultimately I'd like to find a way to get or make solid disc 16 and 20" wheels. This would reduce the number of parts by about 100-140 (spokes and nipples) and elminate the need for adjustment. I speculate (fantasize?) that a solid disc wheel can be made at an acceptable weight, especially if the hub brake is designed to be integral. -- I take responsibility for everything I say, but I could be completely wrong. ============================================================================= Nickolas Hein | Voice: (206) 237-8935 Boeing Commercial Airplane Group | Boeing net: neh3568@aw2.fsl.ca.boeing.com P.O. Box 3707, M/S 96-04 | Seattle, WA 98124-2207 | ============================================================================= From: bilbo@Kodak.COM (Charles Tryon) Date: Fri, 6 Nov 92 9:33:34 EST Subject: Re: Drum brakes (WAS: The Spirit of Human Effort) Nickolas E. Hein writes: >> > ... Ultimately I'd like to find a way to >get or make solid disc 16 and 20" wheels. This would reduce the number >of parts by about 100-140 (spokes and nipples) and elminate the need for >adjustment. I speculate (fantasize?) that a solid disc wheel can be made >at an acceptable weight, especially if the hub brake is designed >to be integral. I'm no expert on wheels (I don't even play one on TV), but I seriously doubt that you will be able to come even close to the cost, weight, performance and robustness of a spoked wheel by using a disk. Spoked wheels have been around for a _long_ time, and _lots_ of alternatives have already been tried. A disk is a lot better when it comes to aero drag, bit they have their disadvantages too, primarily weight and cost. I don't think they are failure proof either (though I've never used one). Talk to Jobst (or r.bicycles) if you want information on wheels. -- Chuck Tryon work: bilbo@bisco.kodak.com B. Baggins home: bilbo@Bah.Rochester.NY.US At Your Service ______________________________________________________________________ Experience is what you get when you were expecting something else. From: calhpvz@bromine.Berkeley.EDU (Eric Anderssen ) Date: Wed, 16 Dec 92 16:12:40 -0800 Subject: hub brakes There are serious defects in the present design of most all hub brakes as they relate to wheel longevity. In fact for a poorly built wheel, the stresses introduced by the torque transer from hub to rim (via the spokes) can lead to catastrophic failure--"potato-chipping, or taco-age." First, look at the best case scenario; that is that both flanges of the hub transfer the torque to their spokes evenly. Here I'll also assume a linear approximation for the induced stress in the spokes. For a dished wheel, the drive spokes have a shallower "diaphragm angle" than does the right side spokes. This means that for a given "wind-up" at the hub (relative to the rim) that the spokes on the left (non-freewheel) side will have a higher lateral force than will the right side spokes. Henceforth I will refer to the sides as drive side and brake side. For a given input torque there will be a resultant force at the rim which wants to push the rim to the brake side, because the spokes on that side have a larger lateral component. This means that when you apply the brake, the rim will move to the brake side until the forces cancel. If both flanges move together (this is the infinitely stiff hub model), then the actual displacement of the rim from center is small--it varies from one to three millimeters. Whic depends on flange diameter, spoking pattern, and dish. I suppose it's important to mention what my load considerations were, so here they are. First, as a ball-park figure, I know that the fastest a bike can stop is about .6g (obtained from "Bicycling Science," Wilson) For a recumbant, with low center of mass, and about a sixty/forty, rear to front weight dist. on the wheels, the weight shift due to the negative acceleration yields roughly a fifty/fifty split. With that info you can find the torque that the wheels must provide-- if you use acceleration as I did, the coeffficient of friction cancels, and you can use your wheel diameter, and rider mass (times .6g) to get the torque. I used ~200lb mass. For a comparison, I wanted to see what kind of torque a human puts out to accelerate-- ie the drive torque. We hold the world record in the standing start 1/4 mile drag, so I figured I use that for the torque profile-- I assumed linear acceleration (which actually would undershoot a little) we did the distance in a little under 27 seconds. Well, the end result is that the brake can provide around six times the torque. It would seem that these forces should be reckoned with. Now, here's the problem I have with most hub brakes: no hub is infinitely stiff--there is always some torsional wind-up in the hub. Of course this can be alleviated, but not too damn many manufacturers, or designers for that matter do it. Allow me to digress. When supporting something with multiple rigid members, generally, a designer has to look at the load paths--namely, which members take how much load. It turns out that the stiffest members take the load. For instance if you support a plate with 100 columns, and make one steel and the rest jell-o the steel one will take most of the load. The jell-o will still take some of it. If the steel is much much stiffer than the rest, say 100 times stiffer than the sum of the rest, then for all intents and purposes, it takes all the load. What this means in the hub system, is that the hub body must be much, much, stiffer than the spoking pattern. Because as soon as the hub starts to wind up just a little, some of the load will be eaten up by the spokes. To make a hub body much much stiffer than the spoking pattern, you have to have a model for the torsional stiffness of the spokes--I'll forgo that here, but for large torsional (angular) deflections, the spoke model gets heinously non-linear. Now, I can begin to explain. For a given torque input from the brake, iff the hub body is not torsionally stiff, even more wind-up will go into the brake-side spokes, which as I pointed out earlier, have a larger horizontal component than the drive side spokes. Now because the wind-up is greater, the forces pushing the rim to the side are even greater. My simple model yielded deflections exceeding 5mm. That is pretty damn far for a rim to move. Say you're braking and either hit a rock, or are initiating a turn--either will put a lateral stress on your wheel possibly just enough to taco it. Additionally, as Mr Brandt writes in "The Bicycle Wheel," Only the five or six bottom-most spokes take up the normal loads off the static weight of the rider. Adding the differential horizontal forces in this region of the wheel is very tricky. I used a linear approximation in the beginning assuminmg that all the spokes were at the same tension--the bottom five are at a lower tension. What I believe is happening in this region is one of my primary reasons or concern. (that should read 'for concern'--my 'f' button is collicky) If you look at the deflection of spokes in a wheel at a lower tension, the lateral deflection is greater. So in the spokes at the bottom, they are shoved to the side just a little farther--GEE!!! doesn't that look like the initial conditions for potato-chip collapse of a wheel!!!!!!!!! In my investigation, several options to alleviate the problem presented themselves. The important thing is to match the lateral forces at the rim for a given input torque. To do this you can do a number of things. Namely have any extremely large hub body diameter--like the phil wood or american classic. Mix spoking patterns per side--this can only alleviate the problem, and is very hard to calculate. BUILD DISHLESS WHEELS!!! You can play around with different flange diameters--higher on the drive side to give that side a larger side component. Devise a way to apply the torque in the middle-- this factors out the need for a torsionally (exceptionally) rigid hub. There is a hub out there which is the embodiment off all of these qualities, but alas, or perhaps gladly, it will not, come with a hub brake. It is the Magic Motorcycle hub/wheel--best call them for the info (206)321-4809. It ain't cheap, but the whole damn thing weighs about 700 grams. It has centrally located pawls and symmetric spoking-- aluminum spokes to boot. Eric Anderssen, UC Berkeley HPV From: pardo@cs.washington.edu Date: Wed, 16 Dec 92 17:10:43 -0800 Subject: Re: hub brakes >[Reasons why hub brakes might not be such a good idea.] All good points. In practice, I have not seen wheel failures on wheels with hub brakes. That tells me that (a) the hub brakes don't weaken the wheels exceptionally (b) not many people use hub brakes (c) those who do aren't typically loading the wheel as heavily as sport-oriented riders. Remember that braking loads are taken partly by the relaxation of forces on the ``pushing'' spokes. Suppose the hub shell is made of layered goose grease and, therefore, has little torsional stiffness. All braking loads are taken by the non-freewheel side spokes. As the brake is applied, half the spokes on one side get greater tension -- and half the spokes get lower tension. Until the slackening spokes go totally slack, there are no induced lateral forces. Hub brakes *can* cause enough load that the ``pushing'' spokes go slack, at which point the lateral forces do become nonzero. I have seen a stiff rim used with a hub shell for a rear wheel (close-spaced large diameter flanges). The wheel was slightly dished and used as a front wheel. A drum brake was screwed on to the hub and used as the bike's primary brake. This seems like a sure recipe for disaster. In 5 years of regular, moderately abusive riding, there were no wheel trueness problems beyond normal pothole bashes. The above wheel replaced an old Sturmey-Archer with 20 years of service and no trueness problems but brake pads were no longer available. I used Sturmey-Archer VS/VT drum hubs with very large flanges as an undished front and dished rear on a single used for mixed commuting, smooth singletrack, and riding down stairs. After 5 years the front wheel needed retensioning. The rear wheel was destroyed by accident. It was rebuilt with a box-section rim and put on a tandem with the drum used as a drag brake. Although it has only 36 spokes, the hub braking load is minor. In 1500kms with light riders, I've had no problems. These are, of course, anectdotal, but they do show that wheel strength with hub brakes is not a first-order concern. ;-D on ( This old hub ) Pardo From: Torsten.Lif@eos.ericsson.se (Torsten Lif) Date: Thu, 17 Dec 92 08:45:07 +0100 Subject: Re: hub brakes Hmm. All seem to assume that a hub brake sits on one side of the hub and introduces torque in the hub body. Maybe there's a reason why the coaster-brake single speeders became so popular here. The entire hub body (from flange to flange) is the drum and the expander sits in the middle, making for more-or-less symmetric distribution of the braking force. With a single sprocket, they're virtually dish-less. I don't know if the coaster-brake 3- or 5-speed (and newly presented 7-speed from both Shimano and Sachs-Huret) hubs have truly symmetric braking, but the hubs are wide (> 2") and should spread the torque quite well. What does everybody have against coaster brakes (except for the fact that they don't work with derailleurs)? It seems to me that adding a third cable (the brake) to a hub that already has two (most 5-speed hubs do) is merely making things more complicated. For the 2-hub-gear design that we've been arguing lately, there will be lots of gear range without any derailleurs, so a coaster brake would work like a charm, or...? /Torsten From: "Nickolas E. Hein" Date: Thu, 17 Dec 92 6:58:45 PST Subject: Re: hub brakes I'd like to add my two cents worth here as well. I've been using the Arai spin-on drum brake on the front of my streamlined commuting recumbent. In about 1000 mi. of real-world riding the haven't given me any problems other than wearing out one front tire faster than I expected. I take this as a good sign that the brakes are working well. I've put a newer and more rugged front tire on and haven't noticed any excessive wear yet. To summarize, I tried them and they worked. All the talk here about spoke concerns or (possible) problems has made me wonder if it wouldn't be worth getting rid of them, expecially for the small (16 and 20" ) wheels I'm using. Would anyone care to speculate on building a custom all-aluminum wheel where the hub and rim are connected by aluminum discs (instead of spokes), maybe with honeycomb between to give buckling support. Although I realize the weight of such a wheel might not appeal to anyone who is strictly racing their designs, it has appeal for a commuting vehicle because it would eliminate lots of parts and, if built true, never need adjusting or replacing. I expect that front suspension will eliminate the need for flexibility in the wheel so its rigidity would not be a deterrent. Just for reference, I currently use a 16 x 1.75 front and 20 x 1.75 rear rim. The front hub is custom made to fit press-in bearings and has freewheel threads on the left side for the spin-on drum brake. The rims are unique in having the same number of spoke holes as hubs made for larger wheels. This was necessary because the spoke holes on the hub were patterned from a conventional hub. The rear hub is conventional (store-bought) and the rear center-pull caliper brake is rarely used because it is so ineffective and obnoxious. From: pardo@cs.washington.edu Date: Thu, 17 Dec 92 14:30:54 -0800 Subject: Re: hub brakes My view: Coaster brakes are popular because they provide years of reliable service without maintainance. Enthusiasts often prefer hand-operated brakes, even in the absence of derailleurs, because they provide better modulation and, in some cases, lower residual drag (many coaster brakes do not roll forward entirely freely, though the drag is small). Utility riders often prefer coaster brakes for simplicity and reliability; hand-operated brakes will often sieze if not lubricated periodically. I have a Sachs 5-speed with a coaster brake and my biggest gripe is that they have provided no lubrication ports and then both claim that the hub is ``permanently lubricated'' and also refer me to ``authorized dealers'' for lubrication. As far as I know there are NO authorized dealers on this continent -- so they are no help. My second biggest complaint is that the hub is clearly less efficient than a well-lubricated derailleur system (and, I think, less efficient than Sturmey-Archer 3-speed hubs), but I am generally willing to sacrafice some efficienncy on a utility bike. ;-D oN ( E' fish an sea ) Pardo From: mason@asylum.sf.ca.us (Latte' Jed) Date: Fri, 18 Dec 92 04:53:17 -0500 (EST) Subject: Re: hub brakes > From: Torsten.Lif@EOS.ERICSSON.SE (Torsten Lif) > > Hmm. All seem to assume that a hub brake sits on one side of the hub > and introduces torque in the hub body. Maybe there's a reason why the I'd actually forgotten that these brakes (that sit on the side of the hub) exist. The ones I'm used to and picture when I heard "drum brake" are the Atom drum hubs, which are an enlarged hub with the brake inside. They're cheap, inexpensive and cheap construction, though they do offer good stopping power and I've never had any problem with them, and I've only used them in harsh environments. The last one I bought cost $12, which is a far cry from a Phil Wood drum brake, and the large diameter flanges combined with a 20" rim make for an incredibly stiff wheel. > What does everybody have against coaster brakes (except for the fact > that they don't work with derailleurs)? Yes, and drums don't offer good control. Drums, coaster and otherwise, put loading on the spokes I'd really rather not have, and when mass producing a bike it's probably cheaper to build in caliper mounts and put a caliper brake and on. From: bilbo@Kodak.COM (Charles Tryon) Date: Fri, 18 Dec 92 9:25:50 EST Subject: Re: hub brakes Latte' Jed writes: > >> What does everybody have against coaster brakes (except for the fact >> that they don't work with derailleurs)? > >Yes, and drums don't offer good control. Drums, coaster and otherwise, > (ect...) Has anyone here had the opportunity to pull apart a real old coaster brake, one which has sat around for a while or perhaps has been overheated? They are filled with the Stickiest Substance Known To Man! They are normally filled with grease, but this begins to harden with time and will really gum things up. After a while this causes the brake to fail, since it won't grab when you back-pedal Another problem with coaster brakes is that they are MUCH more likely to overheat and fail on long hills. Since they have a much smaller area over which to dissipate heat, they can build up enough heat to weld themselves together. (Don't try this at home folks.) -- Chuck Tryon work: bilbo@bisco.kodak.com B. Baggins home: bilbo@Bah.Rochester.NY.US At Your Service ______________________________________________________________________ Experience is what you get when you were expecting something else. From: Eric Schweitzer Date: Fri, 18 Dec 92 12:48:35 EST Subject: Re: hub brakes On Fri, 18 Dec 92 9:25:50 EST Charles Tryon said: > > Has anyone here had the opportunity to pull apart a real old coaster > brake, one which has sat around for a while or perhaps has been > overheated? Many times, for many years. > They are filled with the Stickiest Substance Known To Man! This has NOT been my experience. > They are normally filled with grease, but this begins to harden with > time and will really gum things up. After a while this causes the brake to > fail, since it won't grab when you back-pedal This is often caused by a worn spring clip, but see below. > > Another problem with coaster brakes is that they are MUCH more likely to > overheat and fail on long hills. Since they have a much smaller area > over which to dissipate heat, they can build up enough heat to weld > themselves together. (Don't try this at home folks.) I've never seen this one, either. > As a ritual disclaimer, most of my coaster brake mechanics were done on the north shore of long island NY, a hilly but not mountainous area. I can recall one run or 'sticky' coaster brakes, but as they were all Bendix 76's (Mexico), we figgured it was a bad type of grease used. The problem was not due to design as the later hubs (same model) didn't have the problem. The problem with these (if I recall correctly) was that the grease hardened and the drive clutch would not engage. The brake worked. The only standard brake failures I can recall were due to wear, and only one happend suddenly. Depending on the brake, if a spring clip failed (part of a bendix or centrix design) the brakes would fail. As this was a sudden failure, loss of braking was sudden. In general, on brakes with brake shoes or multi plates (like new departure or NGK) failure due to wear was preceeded by excessive peddel travel before the brakes would engage. This is not to say that such brakes are trouble free, just that when compaired to rim brakes, they require a lot less maintainence and fewer parts. In the cases you've seen that the brake failed from overheeting, how would a rim brake have done? Sure there is smaller serfice area, but the temps required to damage the parts are much higher than needed to melt brake shoe rubber (I think). Just another $.02 from an old mechanic.