Vintage Workshop
Services for Brough Superior motorcycles and their contemporaries

Castle forks reproduction date: 03/2022

When I finally obtained my nice replicated frame, the forks I had made from a set of 1915-ish Harley back legs proved too short for it.
I could possibly have used them with a big spacer under the headstock. This would have meant boring the stem out a second time, making a longer one, lengthen the top end of the tubes somehow and reassembling the whole lot. But as I already wrote some 15 years ago, neither the Castle nor the Harley rear legs have been too strong when new and the ones I had got were not too good either. So I thought if they could be made in 1927 why shouldn't it be possible to make them these days?

A few years ago, I had the misfortune of offering someone to rebuid his seemingly good Castles for an MX100. I thought it would just be a matter of re-tubing the front fork with new spring shrouds, straightening everything a little, repairing the battered damper plates and their push rods and bushing the bottom rockers but in the end the forks turned out to be irrepairable.
When I had the main legs rigged up to check them for straightness I found everything a bit out of line, but not too badly. I usually straighten such things by giving the part in question some pre-load which is still within the range of its elasticity and then gently heat the area where I want it to yield. In doing so I became aware of some strange noises, a faint creaking when I levered on the tube ends.

Further investigation showed that there was a crack in one leg just below the triple plates which had been covered with braze. I have seen such a thing now two more times. If you consider buying an expensive set of Castle forks and it has a thick fillet brazing below the triple plates better steer clear of them - it is probably a patched up crack! You might want to believe (or be made believe) this is only a poor brazing joint that has been re-brazed - you'd better not believe that...

The owner of the forks wanted me to go ahead with the repair, so I investigated further. Dye penetration testing did not look too good, it showed further cracks in the lowermost plate.

So I cut the lowermost plate to better see the extenzt of the damage.

I found that the fork leg had been very badly repaired, with both the inner and the outer tubes being broken. The wole lot had been patched up with the insertion of a pegged piece of bicycle fork tube, apparently in the hope to stabilize the brazing on rusty steel.

Well, to cut a long story short: I had to realise that my focus had been to narrow - when investigating further I found more and more cracks.
This is what remained of the forks, and with the material apparently so tired I refused to go on with the repair.

What has all of this to do with reproducing these forks?

It told me there must be a demand for well reproduced Castle fork back legs and it gave me the opportunity to really see how these forks had been built.

I understand the forks had been made by pushing a shorter length of tapered tubing into the (tapered) main tube, then swaging the upper end down to its relaively small diameter and finally forging it into the S-shape red hot in a die. At which stage they forged it from round to elliptical, or whether they used elliptical tapered tubing I don't know.

Realising that it is very hard to find a manufacturer doing such a thing today I looked into modern production methods such as hydroforming and laser additive manufacturing. The result was sobering: The companies doing hydroforming told me the wall thickness is too great for this method to be applied and the die alone woud cost some 20k.
The laser additive guys said their biggest machines for steel (that was in 2019) have a working space of 500x500x500mm. They said you could possibly do a pair of legs diagonally within this space, but as you pay approx. 5-10k per charge this did not appear too attractive to me... well, maybe there will once be bigger machines so you could do 10 or 20 legs in one charge...

So it was back to the more conventional methods.
My research led me to an Italian company producing "tubi conificati", i.e. tapered tubes. They use tapered dies made from thick-walled tubing into which they push the tube to be tapered, thus reducing its diameter. I thought that would be a good start and skteched a production process, see picture. It became apparent, though, that all the tubes they had made so far were for decorative purposes only, i.e. for floor lamps etc. As 25CrMo4 wouldn't tell them anything and they said their main material was brass I abandoned this appproach as well.

If you browse youtube for "tube swaging machine" or "tube tapering machine" you find things like Especially the last one looks to be just the ticket, I even considered buying such a machine, but trying to make contact to a number of these companies led to nothing.

An especially interesting machine appears to be this one as it is a CNC machine that can produce an arbitrary cross sectional progression without special tooling. I did not want to buy such a machine but tried hard to get some information from their makers. I hoped they could name a customer who had such a machine so I could ask them if they would make my tubes. Well, nothing came of it - I ended up with the impression they had never sold one of these machines...

Eventually my research led me to Paul Brodie, a keen enthusiast in the States. He built a batch of re-created 1919 Excelsior racers, and needed for tubes for the forks. You can have a look at one of his bikes at Yesterdays, the Dutch vintage bike dealers. He got me into contact with a chinese family producing bicycle frames in the US who had made these tubes for him. They look similar to what we need here, but, due to the structure of these "keystone" forks they are of lighter gauge and only single-walled as I understand.
Paul told me the production of the tubes took some two years of experimentation and he had to order a batch of 100 tubes - or was it 100 pairs? Should you need some, I think he (still) sells them for I think 120$ a pair.
I then had a longer correspondence with the Chen family but nothing came of it in the end - ok, maybe I have not tried hard enough.

Eventually my enthusiasm petered out and I resorted to repairing a quite battered fork I obtained in exchange for my efforts in the above affair. It was slightly better than the one shown above but it was a hard job. I'll make a page about the rebuild if and when I find the time.

If one of you out there should feel motivated to go ahead with reproducing those fork legs, please let me know - we might join forces...


Any kind of feedback to is appreciated
(sorry, this is not a clickable 'mailto:' hyperlink. If you want to write me, please type my address in your mailer. )

Back to HOME page - Back to SS100 Replica Project main page