Quality? (round 1)


You know what I mean when I say quality, right? Nope. I mean something completely different.

There are quite a few concepts that are in common use, but are imprecisely defined, or useless as defined. The ideas they represent seem simple - they aren't. The combination is lethal. It's a problem with our language and how we use it. Famous philosopher dudes hassled with it until no one wanted to talk to them about it anymore. I won't go into what they proposed as a solution. It wasn't very practical. What did you expect; they were philosopher dudes, not engineers.

But, exposing fuzzy thinking and bad logic is a diversion that I'm drawn to for some reason. I was born to debunk. It's twisted; I admit it. Trying to restore order to society's collective brain and point out that certain ideas, things that are held as true and useful, aren't really either of these is often not considered a friendly thing to do. But the problems that mistakes in logic and reason create for everybody are much worse. I'll risk it.

There are too many of these fuzzy ideas lurking in our daily conversations to ever hope to get them all sorted out. And, in spite of my innate tendencies, I realize that it's really not my job to do this sort of thing. I'm not a truth cop or a philosopher; I'm just an engineer posing (badly) as a journalist in this case.

It's worth the effort (as a philosophical exercise if nothing else - a scary proposal in a bike magazine, and one that's probably only possible in the UK) to think about this for a second. There are many things that are true and not very useful at all. I've studied enough science to know this. A great deal of modern theoretical physics gets this rap. Seems true, it's interesting if you're bent that way, but it hasn't much practical value. And there are also some things that are absolutely false but can be useful (not always for just purposes). The foundations of racism come to mind immediately (as an extreme example). Propagandists make their living with this stuff. But, occasionally you'll bump into something that is both true and useful. The second law of thermodynamics is both true, to the extent that we can confirm it experimentally, and damned useful. Little escapes its scope. Beer is another, but I'll leave that for now.

Unfortunately, cycling is a rich domain for the not true but useful items, though none are as heinous as the example I mentioned above. We do have more than our share of screwy, corrupt ideas though. Most survive because they work out to be useful in some way. For example, the notion roadies have about dropping a plumb bob down from some obscure bump on their knee to set their horizontal saddle location is on the list of clinkers. This is not a true notion. It's somewhat useful by a quirk of fate, but, it is not true or based on anything rational. When physical arguments are cooked up to explain why it works, and they often are, the apologists are doomed to rationalizing falsehoods, an ancient vocation, but no more respectable now than it was centuries ago.

That's enough ranting about the easy ones; there's no challenge there. It's more interesting to go after the subtle ones. They are sneakier to detect and harder to dislodge. They are familiar, so they are much more deceptive and potentially destructive. The concept of quality is one of these thoroughly mangled intellectual entities.


That's right. Quality. It's preying on your brain cells (and your wallet) constantly.

Confused? Okay, I'll demonstrate what I mean. What do you mean when you say that something is high quality? (Don't cheat, work it out before you read on.)

How do you describe smooth turning cartridge bearings in expensive aftermarket hubs. High quality, right? Perfectly executed welds in spendy ti frames. Premium quality. And the shiny anodized finish on anything and everything milled or turned from aluminum for a bike. High quality again, right?

When the term is used this way there are several things it communicates other than the tactile impressions in the utterer's brain. Subtle things, often implicit, unspoken things. For example: Something that is high quality can cost more. Quality costs, right? It is less risky to buy something that is high quality, since, while it may cost more, it's going to be a better part. It will do a better job because it's high quality, right? The notion of quality helps sell, and it helps sell for more.

But, this all breaks down when you try to convert the ideas associated with the tactile impressions you get twiddling parts into real measures of increased performance, longevity, or durability, attributes that distinguish a part from the rabble, and even justify spending a bit more in some cases. If a tactile impression, leading to a claim of high quality is reliable, then there should be some way to connect it with something real. (Admittedly, if your use of the concept is entirely based on impressions and appearance, none of this is a problem for you, though your wallet has to be based on a substantial bank account for this to work.)

So what about smooth bearings? Where does this go wrong? The smooth cartridge bearings in a hub, off a bike, do not indicate much about the hub's performance for a couple of reasons. Ask any roadie who's been around for a while. When you adjust the cups and cones of an old Campy hub, you allow for some shortening of the axle when the quick release is tightened. If you adjust the bearings so that they were perfect when the wheel was off the bike, they're too tight when the QR was cranked down. Don't believe me? Ask somebody old. Try it yourself. Same goes for a hub with cartridge bearings. The eventual preload on the bearings may be acceptable if you are lucky. Or it may not and the bearings will die quickly. But you can't tell in advance while twiddling the hub. The smooth feel seduced you. It's a false promise.

And a smooth feeling hub (again, before it's built into a wheel) may have another problem. The way a radial contact cartridge bearing fits into the hub determines how much internal clearance or preload it has. Too much press and the bearing is tight. Too little and it's loose. When you build a wheel, spoke tension stretches the hub flanges radially outward a bit, especially if the wheel is built radially. The bearing bore stretches open. The bearing's clearance does too. You can't tell what's going to happen before you build the wheel (actually you can, but it depends on how you arrange the spokes - way too complicated). So the feel of the bearings before the wheel is built is not a good indicator of how they will work afterwards. Smooth bearings are smooth bearings, they do not necessarily indicate a good hub. Funny, eh?

What about the appearance of a ti weld? The concept of quality doesn' work well in this case either. There are a lot of ways to make a marginal weld look good. It's part of the art. Welders learn how to do it. I know how to do it. It is not a big deal. (Don't worry. It's rare that a truly fatal weld is covered up with a cosmetic pass. That's not what I'm talking about here).

There is very little connection between the smooth, regular appearance of a weld, in any material, and the performance or durability of a frame. In fact, given that a straight frame with minimal residual stress often is best achieved using a weld sequence that requires running beads of relatively short length, alternating about the center plane of the frame to achieve (almost) mirror symmetry in the distortion the welding creates, a slightly irregular weld appearance may be optimal. A smooth weld is not a deception if it is represented as nothing more (or less) than a smooth weld. It is a deception if it is used to demonstrate anything else about the frame.

And most cyclists have seen enough shiny machined aluminum bits in the recycling bin before their time to know that there are silk purses converted from sows ears with a little polishing and some time in the anodizing tank. Not all of them are; don't get me wrong (or beat up at the next big trade show by shiny aluminum parts manufacturers). A part with a shiny finish is a part with a shiny finish, nothing more. The finish can even detract from performance.

So, in order to sort this all out, I'll propose a new (to the cycling public, at least) definition for quality, one that is much more restrictive in the range of instances in which it can be used properly, but one that is consistent and useful. Quality is a measure of the extent to which a part conforms to its specifications. Want to talk about bearings? Talk about bearings. Talk about bearings in terms that bearing engineers do, the terms that describe the way bearings work or don't work. Talk about the bearing's specifications. Load ratings. Seal materials. Fits. But don't use the word quality. Same with ti welds. Same with aluminum bits. Use terms that are directly connected to the parts specification or function.

All the other uses of the term are deceptive.


Quality? (round 2)

I went after a concept many of us use when we talk about bikes last time around - quality. I tried to show that the way the term is most commonly used isn't very useful. It can be misleading and confusing. It often doesn't mean what we think it does. That's a problem.

This may seem like a waste of time to you. What difference does it make whether we call it quality or something else? KB's just going off about some technicality of the language. I don't think so.

My motivation is simple. I think mountain bikes are technically complex mechanical devices that have a fashion element to them. Of course, you can look at it the other way around as well (Isn't that right Chipps?). Mountain bikes can be thought of as rolling fashion statements that happen to be something that you can ride around on. If you believe the latter statement is true, I'm wasting your time. You have more pressing things to ponder, like which color of anodized aluminum bits you are going to spend your food money on this month. If you think the first description was the accurate one, the thing matters to you.

You are a mountain bike part consumer. You have to pay to play this sport, whether you want to or not (unless you are in a position to get free stuff, right Chipps?). How do you decide which bike to buy? If you ride hard you break things or wear them out. How do you decide which new bits are the best replacements for the old stuff? If you're a fashion hound, it's easy. Follow your instincts and your eye (and your nose if you're really weird). But if you think a bike is to be ridden, you want the parts to do a good job, and you have to think a bit more. A screwy concept of quality messes this up. Stuff that should be looked at hard gets the green light to buy because of a poor use of the concept. It's your gold that's at stake.

The important characteristics of mechanical devices can be identified, and they can often be measured. These characteristics describe the way the part works in the field. Is it a good part or not? Does it help you ride better or not? What do you need to know about the part to figure this out? Its mechanical characteristics. I'm talking about things like durability (of course), frictional properties (for moving bits), stiffness (for a few things), weight (okay, it matters), and a few other things (depending on the part and what it's supposed to do).

Last month I tried to show that I don't think it's logical or useful to think of a part's quality in the common way. An axle that turns smoothly in a hub, determined by the notoriously precise "hold it in your hand and spin it" method, is a pretty useless measure of bearing performance. If you want to know whether a part is good, what you should do is learn something tangible about it. Figure out if it's a good part. Is it designed to perform well? In what way? Is it's specification well thought out? What about the spec works for you? Or is it that you really like the smooth polished finish that wastes so much of your hard earned coin but looks so nice and is so stimulating to the opposite sex? It's the part's design and specifications that determine whether it will perform well; but it's often the external finish that seduces you into buying it. You're better off recognizing this right up front. Don't just muddle it up and call it quality. If you want to get the right bits, you have to get away from imprecise characterization of parts. It's your gold.

I have one more comment about quality while I'm on this rant. It is commonly held that anything made in China or Taiwan is of suspect quality (less so in the case of Taiwanese products these days), while the stuff that comes out of the US or Japan is generally of good quality, right? (Okay... if it comes from Yorkshire, it's pretty damn good stuff too).

This is lame. The concept of quality is used the wrong way again. It's the comparison between the performance and value of the parts made in each place that's important to most folks.

To fix the thing, exchange the term "performance" for "quality" in the comparison above, because it's really the design and specification we're thinking about. The comparison then reads: the performance of parts made in Taiwan and China is worse than that of parts made in the US, Japan, or the north of England.

Does the paragraph above mean that the manufacture of good performance parts or bikes is a racial tendency or a geographical phenomenon? This might be what is intended sometimes. I sure hope not.

If a high performance specification was written for a given part and it was designed right, and the production parts always conform to the specification, it is, by my definition, a high performance part made to good quality standards. It doesn't matter where it is made or what language the people speak who make it. It's clear to me that high performance parts can be made to specification in Taiwan and even China. These people can also out-design us and do. So why do we believe that they don't make high quality parts?

First of all, it's because the statement about the relative performance of parts made in Taiwan and China, and the US, Japan, and the bitterly cold north of England is true on the average. If a part's specification and design are written primarily with low cost in mind, the part will not fare well in a comparison with parts designed with performance as the highest priority and cost a secondary consideration. It can't. If you design with low cost in mind you have to compromise something. The part will weigh more because it is made with less expensive, less durable materials. It will have less finish work done on it because finish work is expensive. It will not be made to exceptionally tight tolerances, only enough to function properly (hopefully). It's performance can often get close to that of expensive parts, close enough that it won't matter much to a lot of people. It won't be as good in every respect. And it may be a quality part if each one conforms to the original specification - even if it's performance is not great.

It turns out that, in my experience, many foreign suppliers are capable of delivering well designed parts that conform to high performance specifications. They can if they are asked to do so and if they are paid to do so. But they rarely are. And there is a sound reason for this: companies that make parts in China and Taiwan work hard to keep the prices of their products reasonable. You can't build a $500 mountain bike if you get carried away with high performance specifications. You have to contain the costs of every component, rigorously, or you quickly end up with an $800 OR $1800 bike. This is a tough job.

This is the part of the business in which the Chinese and Taiwanese bike companies make their livings. The parts they make are often not designed to give exceptionally high performance. They work, generally well enough for their intended use. They frequently conform to the specifications, though this is a much deeper subject than I can deal with now. Comprehensive, meaningful specifications are not simple to write and are not common in the bike biz in my experience. Aspects of a design that are not thoroughly specified can cause problems, and often do. But they can do it.

So you can't determine the performance of a part based on where it was made. It's never easy is it?