Unravelling the myths, half truths, and old wives tales that surround "Framebuilding"
by Dave Yates
Greek myths and legends have nothing on those surrounding the design and construction of bicycle frames. I have been building frames for over 30 years now and I have seen or heard most of them! There is nothing mystic about a collection of eleven tubes, usually of metal, stuck together with some form of "glue", be it hot metal or cold adhesive. In its simplest form a bicycle frame is a utility device to hold the parts of the bicycle in a working relationship that has changed little in the last hundred years or so despite many attempts to reinvent it. With all due respect to our recumbent riding colleagues the greater proportion of all bikes on the road are conventional "diamond " frames. This is what I have made or designed over 11,500 of and this is what I intend to concentrate on, unravelling the myths, half truths, and old wives tales that surround "Framebuilding"
First, a little about myself. I am self-taught. I was for the first ten years of my working life a teacher of technical subjects, mainly metalwork, in a Newcastle Comprehensive school. A passion for cycling and an insatiable desire to make things inevitably led to messing about with bike bits in general and frames in particular. Having had a go at various frame modifications and repairs I eventually thought it can’t be that difficult to build a complete frame, so I did. (Although I did have the advantage of starting off with a good grasp of the necessary skills and a rather fine workshop to practice them in!) At first I simply copied and adapted, it was not too difficult to work out what worked and what was impractical. By riding my first creations I was able to refine the design aspect by trying it out myself then on my clubmates who were queuing up after the first two or three seemed to work OK! After a couple of years I decided to pack in teaching and attempt to make a living at building frames, reasoning that if it did not work I could always go back to teaching. That was thirty years ago and the rest, as they say, is history.
There are two aspects to creating a frame that the rider feels completely at home on. The first is the "plumbing" i.e. the builder must have the skill to join the "pipes" together securely and in the correct alignment so that the rear wheel follows the front wheel in a straight line and the bottom bracket axis is square in both planes. This is a manual skill that can be learned. For someone with good hand / eye co-ordination and the right attitude, the basics of brazing and welding can be picked up fairly quickly. To get really good takes a lot longer!
The second and most difficult aspect is designing the frame to suit the riders intended use and style so that when they get on it it either disappears beneath them or they think, "this is better than sex". Every builder seems to have his own way of achieving this. Some use extensive computer programmes, often written by a third party. I tend to favour a much more subjective approach. I talk at length to the customer and look at him / her sitting on their existing bike, watch them riding it and generally try and get a picture of their vision of the new bike. The design evolves out of the discussion and observations. There are certain dimensions on a frame that are important. In order of importance by my reckoning they are:
Top Tube /seat tube lengths
Head angle / fork rake
Bottom bracket height / stand over height
Front centres (wheel axle to bottom bracket)
Equally there are others that are irrelevant and are a result of getting the important ones right.
The two numbers I get asked about most are the least relevant in terms of performance. These are weight and wheelbase. Each is a resultant of several far more important design considerations. Every important dimension on a frame is a trade off. By altering one, several others are affected. The trick is to balance all of these into a set of dimensions that fulfils the requirements of the first sentence of this section. Certain standard dimensions also influence the design of the frame. Wheel and tyre size, bottom bracket width, hub width front and rear, and brake type are the major ones. Wheel size is probably the most significant with tyre availability also part of the equation. It is no good choosing what appears to be the ideal wheel size to find out that tyre availability in any sort of quality is non existent.
Whenever I design a frame for a customer the first question I always ask is "What are you going to use the bike for". Each type of use has its own requirements from round the world touring to track racing. The beauty of a custom frame is that it can be fine-tuned to suit the requirements of the rider. I regularly get asked to build frames that can be used for more than one purpose. This is usually possible by juggling the design and material selection. One of the myths is that tubes come in sets and should not be mixed. This is rubbish, I have been mixing and matching tubes for almost as long as I have been building frames. Each tube in a frame has certain requirements upon it dictated by the design of the frame and the use. If the best tube for the job is of a particular material from a particular manufacturer then there is no need for any of the other tubes in the frame to be from the same source. I have made frames before using tubes from four different manufacturers. One I can recall is more than twenty years old and the owner keeps assuring me it’s the best frame he’s ever had.
Lets look at some of the numbers. "Angles" is a much misunderstood and abused term. The head and seat angles are completely independent of one another. The seat angle is the angle formed by the seat tube and the horizontal and is used to determine the fore and aft position of the saddle relative to the bottom bracket centre. It is a function of rider’s thigh length and the use the bike is to be put to. The object of the exercise is to get the rider’s knee positioned above the pedal spindle. For general riding and Audax bikes I normally work on a point about one inch behind the riders kneecap directly above the pedal spindle when the crank is horizontal. For full blooded touring a touch further back, for short distance road racing and crits a touch further forward and for time trials with tri bars further forward again. However this is also influenced by rider preference and riding style. I always aim to get the saddle in the middle of its fore and aft adjustment. This can be quite difficult, as there are so many different saddles on the market with wildly differing designs. I usually work on a Rolls type pattern unless told otherwise. The Head angle is the angle formed between the steering axis and the horizontal. This is combined with the fork offset (rake) to produce the "trail" (see diagram) which governs the handling to a great extent. Weight distribution is the other factor here. The trail is the distance between the steering axis and the point of contact with the ground. It gives the "castor" effect that keeps the bike tracking in a straight line. It is possible to ride just about any combination but the right set up makes the frame relaxing to ride and not a fight to hold in a straight line. I once rode a small wheeled bike that a customer had "modified". He had put so much bend in the fork blades that it actually had negative trail and, unrestrained, the front wheel kept trying to reverse itself. The bike was rideable as long as you kept both hands very firmly on the bars and did not attempt "look mum, no hands". If the builder gets this particular combination right then the bike will ride "eyes shut, hands off" which could be a major requirement for an Audax bike!
Top tube length, which is arguably more important than seat tube length, is governed mainly by the riders torso length and is influenced by the need to keep front centres long enough for toes to clear wheels mudguards etc. I would usually make an Audax bike and a racing bike for the same rider with different length top tubes and as a consequence different length stems. The Audax bike needs longer front centres to allow toe clearance for mudguards and usually slightly larger tyres. A road race bike with no mudguards and skinny tyres can be shorter at the front; indeed this is a good thing for a racing frame, which needs to be as stiff as possible to transmit power. Shorter tubes = less flex. On the race bike the stem needs to be longer to keep the reach sensible.
Seat tube length is not as important as is believed. From an aesthetic point of view, if you want a "classic" looking bike with a horizontal top tube and just the "right" amount of seat pin showing then the proportions need to be carefully controlled. The need to specify to within 1.5mm as I have had before is a touch OTT. If you consider compact style frames with shorter seat tube and sloping top tube, then clearly the seat tube length is a matter for discussion. Saddle height on the other hand is very important. I always work on inside leg plus 9% from face of pedal to top of saddle as a starting point for this then adjust in the light of shoe size, shoe type and pedalling style. Once the saddle height is sorted, the seat tube length can be decided. One factor to be considered is Handlebar height. If the frame is on the small side and you want your bars quite high then the stem will be near its maximum extension. For Audax and touring bikes I always err towards a slightly longer seat tube than for a racing bike because of the need to keep the front end high enough. Another way around this is to extend the head tube or steering column.
The seat tube and top tube lengths are independent of one another. A common misconception is that if the seat tube is "x" then the top tube has to be "y". This is completely untrue, for a given seat tube length the top tube can be any size within the constraints of tube availability and front centre requirements and is usually a function of torso length.
So what happens when you start moving things around? First thing to realise is that the down tube and the seat stays are completely non-critical lengths as far as the design parameters go. Each simply connects two points on the triangle that have had their positions determined by the important design considerations. So if you lengthen the top tube and keep all other important dimensions the same, the front centres increase and as a consequence the connecting length of the down tube increases. If you increase the seat angle and keep the top tube length the same the same thing happens. On the other hand if you want to keep the front end in the same position relative to the bottom bracket and have a steeper seat angle then you can shorten the top tube to compensate for the increased seat angle. The reverse is also true. Raising or lowering the bottom bracket height affects the stand over height and the length of the head tube.
Problems start to arise with small and large frames. For a bike with a horizontal top tube and a sensible stand over height there is a limit of about 19.5" for a lugged frame. The diagram illustrates the problem. If you keep the bottom bracket height at say 10.5" and reduce the frame size by sliding the top tube down the seat and head tubes the limit is reached when the top tube reaches the down tube. The position of the down tube is fixed by wheel / tyre size, brake depth / clearance, fork crown and lower headset stack height. From this point the only way to make the seat tube shorter is to either a) raise the bottom bracket, or b) slope the top tube down to the rear. Another way of getting a lower stand over height is to use shorter cranks to allow a lower bottom bracket height and smaller wheels. This always raises availability / cost factors so I only do this in cases of no other option. The sloping top tube is usually the easiest option. It used to be considered essential that the top tube was horizontal. Fortunately the mountain bike has changed peoples perspective and attitudes and this, thankfully, is no longer the case. To get the best out of this sort of design the tubes must be joined by a lugless method of joining where the joint angles are not constrained by the availability of lugs. Fillet brazing is one such method, expensive due to the hours of hand finishing necessary, but a work of art if done properly. Another is TIG welding; less finishing to do so cheaper but looks like all the other TIG welded frames to come out of Taiwan. You pays your money etc.
At the other extreme the largest frame that can be made from a standard seat tube is around 25.5". Reynolds used to do a "long" seat tube which will go up to 32". The biggest frame we have ever made was 31" for a very tall 16-year-old! The main problem with big frames is to stop them flopping about. Oversize tubes are the answer here along with tricks like dropping the top tube to reduce the distance between the bottom bracket and seat cluster. A compact style frame with a sloping top tube and a longer seat pin is another option. Whatever path is chosen one fact remains. Big frames are heavier because there is more metal in them!
Weight is almost a fetish with some people. They will sacrifice reliability on the altar of the goddess "grams". One of my favourite sayings is "To finish first, first you must finish". I would rather build a frame that is slightly heavier with all the extra metal in the right places and hence more reliable than a mega light creation which may or may not survive. A fact that escapes most people is that all steel weighs the same. If you take samples of several different types of steel in equal sizes they will all weigh the same. The way to make a frame lighter is to use less! i.e. thinner tubes of higher quality steel. It is possible to build a frame weighing under 3.5lbs from available steel tube. A "sensible" Audax frame will usually weigh 4 – 5 lbs depending on size and material selection. The weight is completely dictated by the tubes and components selected and should never be an overriding factor in their choice. A heavy rider who has a less than smooth style and is going to use the bike for touring as well as Audax requires a rather more robust construction than the smooth greyhound type of rider. This is where the builder’s experience and judgement plays a great part in the creation of a successful frame. My 653 frame that I used for PBP weighs about 4.25lbs, the complete bike weighs about 24lbs with no extra bits on. Then I put two bottles on weighing about 3lbs, my survival kit about 2lbs, lighting kit incl. Schmidt hub etc. I have never weighed it loaded, I would probably scare myself silly, but it must be at least 32lbs. The last thing that goes on it is me, all 173lbs of me! This year’s R&D project is a light bike for me. I got it down to just over 21lbs using sensible equipment, Ultegra mainly, and a compact frame. I could have got it much lighter but like Hi Fi the cost starts going up exponentially for very little gain. I saved about 2lbs! The rather long-winded point is that the difference between a "light" and a "heavy" custom steel frame is about 1.5lbs max. In the context of our usage this is inconsequential and should not be taken as a design requirement.
An article like this can never hope to cover the whole subject in any detail but I hope I have been able to give at least an idea of what goes into a custom frame. For any one contemplating the purchase of such a beast the best advice I can give is to seek out a builder who listens to your ideas, preferably by recommendation and draw on his experience. Then enjoy the unique experience of riding a custom frame created just for you.
If you are feeling inspired and fancy designing and making your own custom frame click here
Dave has also written articles on Audax for Cycling Weekly which you can read by clicking here