Bending Bigger Tubes

My plan was to build a bike inspired by the old red Pashleys ridden by the postmen of my childhood. However, these bikes were quite rightly angular, utilitarian and indestructible. I needed a touch more, so I decided to give the bike some deco lines, shaping the whole frame around two flattening curves. The first starts at the rear dropout, travelling up the seat stay, through the top tube and finishing at the front of the rack and headlamp. The second, more simply, went from the bottom bracket to the tip of the rack stays.

The lengths and angles of the chain stays and head & seat tubes were already defined from a physical fitting. These were then drawn out full size and the curves were then chosen and described by a thin uniform strip of timber.

The individual sections of each curve were then transferred onto 70mm thick blocks of birch ply. Again off-cuts, previously glued up together to give the required width. The curves were then cut, sanded and spindle moulded to give each one the appropriate half tube profile.

This jig was not the most streamlined thing I have ever constructed, but it needed enough leverage to comfortably bend a down tube. Metal is the obvious way to go if thinking long term. I was at the trial and error stage here, the learning stage and if the wood is free...

Two metre oak handles held the other half of the former, pivoting on the M10 bolt shown below. The top halves were buffed and waxed while the bottom were left raw for added friction and then the beginning of the tube was clamped down firmly.

I was delighted by how easy it was to use with one person on the handle and a second gently putting weight on the end of the tube to ease it into place.

The next hurdle was the spring-back. Obviously, steel is elastic in nature and will not hold any bend as described unless, of course, you destroy it by taking it too far. Laying the tube on the drawing and measuring the difference between the straight tube and the bend it was momentarily bend to, I calculated that the spring-back was almost exactly 40%.

It did not matter where on the gradually changing curve the measurements were taken, the figure was always the same.

So, the curve was tightened on the drawing by 40% all along, transferred, re-machined and the tube re-bent.

I didn't expect it to be perfect, but it was. It was spot on!

Just as an experiment, I had a go at bending some butted tube. I had a pretty strong idea how this was going to end, but I needed to see.

Little surprise, the moment the arm reached the necking point were the wall thickness changes from 0.9mm to 0.7mm, instant collapse.

Here is the curved block former for the down tube with the 40% extra bend added. You can see the bottom of the head tube in the top right hand corner and the lines of the desired curve and the original straight tube both coming from the bottom bracket, bottom left.

For safety sake and economy of materials rather than time, I followed the cut, bend, measure, re-cut, re-bend procedure for the down tube and the seat stays too. I needn't have bothered, it was always the same, 40%.

The only change I needed to make was to step the seat stay former block from 12.5mm to 16mm over the appropriate length so the stays would sit snuggly and be prevented from flattening at all.

Here are the final curves roughly laid out...

...and here, cut to length and mitred, but pre torch...

...and then after the event. I really liked it.