Wednesday, 31 August 2011

Lower Fork Legs and Lasers

What with the sudden change to fantastic weather and an unplanned encounter with a surgeon's knife, I've spent relatively little time in the shed this week. However I can report on some progress

Last week I mentioned that I would start to look at the lower fork legs. These start at the same elliptical profile as the upper fork legs and taper to a round profile at the other end. Initially I had thought about starting with a round tube, tapering it and then squishing it to the ellipse. Companies exist that specialise in tapering tubing using dies that cold work the tube to the required dimensions, however there are none in NZ that have the dies for the dimensions I need. With the high NZ dollar, it should be possible to have this done overseas and shipped here. I contacted several companies who upon discovering that I didn't require several thousand units seemed to lose my email address.

Plan B is to start with a round tube, squish it in my rollers to an ellipse and then split and weld the join. By simple calculations, it is easy to work out how much material to remove so that when the split is closed the small end can be formed into a circle of the correct dimension.

Some years ago I took a welding course at night school and thoroughly enjoyed it, however I'm out of practise and don't currently have any gear so I'll get somebody else to weld the splits for me. 4130 chromoly tubing can be either gas welded or TIGed using the appropiate filler rod. There are a couple of places in Christchurch that make racing car roll cages out of 4130 so I'll approach these when I get back on my feet next week.

The bearing housings that connect the upper and lower fork legs are machined from 0.7" steel plate. To speed up the process I'm going to get the blanks laser cut from 20mm plate. Fortunately We have a local company that is still operating after the earthquakes that has the capability to do this. They require 2D .dwg or .dxf files to drive the laser so I've spent a ridiculous amount of time this week playing with free CAD packages from the internet. There is a large range to choose from, some are really good and some are not. A lot of the good ones are locked down to only output parts and assemblies in their own specific format and a lot of the ones that do output the standard formats are a bit crap. Many years ago I worked for PTC (ProEngineer) and then later SolidWorks writing their 3D CAD software so I have half an idea about this stuff and I could get up to speed quite quickly. You wouldn't believe how few product are able to produce a fillet between a straight edge and a circular one. For 3D modelling I've been using AutoDesk 123D and for the 2D drawing I've been impressed with DraftSight from Dassault Systemes. I think that with the time I've spent playing with CAD this week I could probably have cut the parts from solid using a nail file. It was fun to relearn some CAD skills though.

In other news, I've finished the month with 1001km cycled which I'm quite pleased with since I lost a week due to snow and ice. I'm now going to be off the bike for weeks which I'm really not looking forward to. Cycling keeps me sane.

Monday, 22 August 2011

Elliptical Tube Rolling

    Unlike an ordinary penny farthing [1], a geared facile has untapered upper fork legs of an elliptical section. I have previously made tapered forks for a penny farthing by slicing two sections from a much larger diameter tube and welding down either edge. If done carefully this is an entirely satisfactory method and produces the sharper edge common on earlier bicycles. More prone to cracking where the head tangs insert into the legs though, a very high stress riser point. Since all the tubes on a geared facile, except the handle bars, are elliptical I need a way to reliably produce elliptical tubing of predetermined dimensions.

Ellipses are tricky things. Very elegant mathematically though.

For example, Everbody knows how to calculate the circumference of a circle or a tube but the calculation for the circumference of an ellipse is non trivial. Fortunately I am unable to accurately use a hacksaw to the 4th decimal place so a common approximation will suffice. I have been using this online calculator. Even though it has bugs.

The front upper fork legs have a major dimension of 1.6" and a minor dimension of 0.7" [2]. This results in an approximate circumference of 3.88". In theory then, if I start with an initial tube of diameter 3.88/Pi and squish it, I should end up with an elliptical tube of the correct dimensions. 3.88"/Pi is 1.235" which is very close to the standard tube size of an inch and a quarter. Now, New Zealand is a metric country and I'm unable to buy 1 1/4" tubing but I can get 31.75mm tubing quite easily. I got mine from here, which is an easy bike ride from my house.

    Tube wall thickness is another consideration since the tube must be able to take my weight without bending or fatiguing but must be thin enough to be squishable and light enought to provide a decent ride quality. The wall thickness was the one dimension I was unable to measure on the original so I referred to my books instead. On my original racing penny the tubing is 0.028" (22 bwg ~ 0.7mm) but this is fragile and provides a very whippy, exciting ride.

Paul N. Hasluck writing in 1897 describes how to make a front driver and recommends the dimensions as 18 bwg for a 'normal' 13 stone man (83kg) and 20 bwg for a light weight racer under a lighter man. 18 birmingham wire gauge is 0.049" and 20 bwg is 0.035". I obtained samples from 0.028" up to 0.095" (~ 0.7mm - 2.4mm) and duly squished them. I have chosen 0.049 as per Mr. Hasluck's recommendation for the more sturdy rider.

So how to accurately squish this 1.25" x 0.049" chromoly tubing?

Some years ago I bought a machine designed to slice thick leather straps into two thinner leather straps, I have no use for thinner leather straps but it was cheap, very well made and I knew I could use it for one of those 'future projects'. The machine has two rollers for gripping the leather and a geared down manual handle for winding the leather between the rollers. The gap between the rollers is adjustable.

I made a set of replacement aluminium rollers with bronze bushes, each roller has half of the desired elliptical profile cut into it. By starting the rollers far apart and gradually decreasing the gap I figured I could roll the tubes to the correct profile.

 In practise this worked very well if I pre squished the tube slightly so that the rollers had a slightly wider contact patch to grip on to start with. The tubing was pre squished by marking a line down the tube and squishing it in sections in a big vice. As soon as the rollers would grip I starting the rolling process, occasionally taking the tube out and turning it over to reduce bending. Since the desired ellipse has high eccentricty (very flat) it meant a lot of squishing, this proved to be very hard physical work.

Chromoly is extremely tough steel to manipulate and twice I needed to have a sit down and a nice cup of tea during the process. The forces involved here are high as I found out to the detriment of my left thumb nail. As the tube begins to conform to the shape of the rollers it begins to produce a pleasing low whistle as it is rolled, this increases in pitch as the end of the tube is approached. However, it is important not to roll to the extreme end of the tube as it can (and will) fly out as it goes past that point where it is no longer contained by the rollers. I don't need a left thumb nail anyway.

There is a small amount of spring back in the final shape but the the resultant tube conforms very closely to the machined profile. There is also a slight tendency for the tubing to twist as it is rolled, the pre squishing mitigates this to a certain extent and as the fork legs are only 21" long the error is managable.

I'll look at the lower fork legs next which start elliptical and taper to a smaller circular circumference.

In other news, now that the snow has gone I'll have to go to work all week instead of getting in some quality shed time. Sigh.

[1] Pun intended. I rather think I'm going to refer to the standard design ordinary bicycle as a penny farthing for the remainder of this blog. When the PF was the standard design it was simply knowen as a bicycle and only later when the new fangled safety bicycles came along was it given the moniker 'ordinary' to distinguish it from the usurpers. The are people who object to the name penny farthing and feel it is derogatory. I personally don't care either way, I use penny farthing because it is commonly understood and less pompous. The American term high wheel also has merit but is not commonly understood outside of the states. The german and french equivalents, hochrad and grand bi resp. are more or less translations of the american term.

[2] This ellipse has an eccentricity of 0.899, which is high. I have nothing to say about that.

Edit - more elliptical tube tube rolling adventures here and here

Wednesday, 17 August 2011

An accidental collector

I'm still not entirely sure how it happened. I used to be a normal cyclist and then somebody offered me an old bike. That much I do remember, I still have the bike and I still ride it. You know how it goes, a few years pass by and before you know it you're dressing in Victorian cycling clothes and embarrassing your children.

My collecting habits are simple, red Italian racing bikes with Campagnolo equipment, red English racing bicycles with Sturmey Archer gears and anything nickel plated. The red thing is also accidental. After a period of time I discovered that I had mostly red bikes. I seem to like red bicycles. As an added bonus I realised that to the layman (my wife) a red racing bicycle looks much like any other red racing bicycle. This made future acquisitions much easier. Also, I do that thing that she does with new shoes. She buys them and hides them in the wardrobe for a few weeks. She wears them and I (sometimes) spot them and ask if they are new. "Oh no, I've had them some time now". It's relatively easy to hide a red bicycle amongst others of it's kind, I just can't ride it for a while.

The geared facile is a happy marriage of epicyclic gearing and Victorian engineering at it's very best, it is a complex machine and in it's day (1887-1892) was rated very highly. I've been keen on getting one for a long time but with very few surviving, availability becomes an issue. During our 2007 penny farthing tour, I mentioned this interest to a friend of mine, a well known collector from Ireland. He let me rabbit on for ages before telling me that he had one. Prior to this I only knew of  7 examples and had personally seen 3 of them. I've since found out about quite a few more but I still suspect that the worldwide sum total is less than 30. I found myself in Dublin in 2008 and spent 2 whole days dismantling, photographing and measuring the geared facile. It is a complete and unrestored example from about 1890. The unrestored bit is important, since copying a restoration is problematic as you don't know what is and isn't original spec. Over the years I've also collected a lot of literature and specs about these wonderful little machines.

Now, if your shed is anything like mine it will house an increasingly large collection of 'future projects'. Stuff that you may or may not ever get around to doing anything about. The facile project has always been on the back burner and it wasn't until I finished my wife's 1904 Royal Enfield that I was able to start, That was about  a month ago. She actually rides it too, I often come home and find it down from the hook.

My wife's 1904 Royal Enfield about a month ago

So I'm about a month into the project and have been badgered into doing a blog about it by Mr. Middleton. I suffer from an acute lack of time, so updates will be as and when I do stuff. Actual cycling trumps everything of course which further reduces the time available.

Right, I'm off to count my wife's shoes...