Wednesday, August 29, 2007

OOPs

Should posted this before the last posting - but I forgot to upload it.

What’s new?

I’m recently back from a family trip to the Eastern Sierras around Mammoth Lakes, CA. It’s a lovely part of the country and a great place to visit. With my lungs, however, I could never live there. We took 5 days to acclimate to the base altitude (8000’). This got me to the point of being comfortable, as long as there were no stairs to be seen. Then we packed in to camp for a few nights at 11,500’. We were on the shore of a beautiful blue lake rimmed with mountains – the sights couldn’t be beat. But, walking up from the lake to our camp (maybe 25’ of elevation) was enough to wind me. Ugh.

Many projects have been on hold for a while, but I’m beginning to finish them. The path racer is ready for paint except for some final filing around the track forks (dropouts). I’m pleased with how it looks, and it should be in the painter’s hands within a week. Janet’s randonneur frame is done and will go to paint at the same time. Charlie’s carbon frame is designed now. See picture below. It will be carbon (from an old trek 2500) w/ a steel bottom bracket shell, chainstays, dropouts and front fork. Charlie wants Gordon to paint this to look like my bike – which sounds like fun to me.

On Charlie’s bike, the BB will be lugged so that the down and seat tubes can be bonded in. This will involve filet brazing and I’ll try to get some pictures posted of the process. The head tube and seat cluster will be wet wrapped carbon that is vacuum-bagged to pull out excess epoxy. The seat cluster will be complicated because I’m using traditional stays (carbon of course) on either side of the seat tube. So four tubes come together in one place and need to be wet wrapped, then wrapped with a material through which the epoxy can pass, but which will not stick to the finished work. All of this then gets wrapped in a bleeder felt which allows air to be pulled out by the vacuum pump and which sucks up and holds the excess epoxy. Then, a plastic ‘bag” has to be put around all of this and sealed up (especially where the tubes pass through) so that the bleeder felt can do its job while the epoxy cures.

For the initial connection of the stays to the seat tube, I’ll be using my old reliable Scotchweld 420. This sets up in about 20 minutes, and should hold things together through the process of wrapping the joint up. A balsa plug will be located in the rear of the top tube so that the top tube can be similarly affixed to the seat tube before wrapping and bonding. All in all, I expect this process to take about an hour from first gluing until the bag is being vacuumed. This setup will be left to cure for 36 hours, then the vacuum is shut off, the bag removed, and the felt and transmission layer peeled off. Using a Mylar transmission layer (generally with perforations) can yield a very nice smooth looking outer surface. Given the complexity of the seat cluster, I’m not planning to use Mylar. Instead a specially treated polyester cloth (woven) will better follow the contours of the tubes. Even then, I expect to see some ridges, creases, and thick spots in the epoxy. My goal will be to sand these down to something presentable before shipping the frame off for paint.

Assuming that Charlie’s bike goes well, I’ll build a full scale prototype for me to ride. This will continue to use steel chain stays as an economy move. I have plenty of experience bonding carbon rear triangles – so I’m not worried about how that works. With two finished bikes in the bag with lugged bottom brackets and vacuum bagged head tubes/seat clusters, it should be time to apply this method to a production bike. I may apply this to Tim’s prize bike. It is a small frame size – which presents a builder with certain challenges in both design and construction. This system takes me away from the constraints of pre-built lugs which will make Tim’s bike easier to get right.

With Tim’s bike, we may also go with an aluminum bottom bracket. Bottom brackets are made in both aluminum and steel with a lug for bonding in a carbon seat stay. This would require outsourcing the TIG work of putting lugs socket son the bottom bracket for the seat and down tubes, but should reduce the overall frame weight. Similarly, I may look at milling grooves in the aluminum head tube. Because it will be carbon wrapped, it may be possible to take a little weight off of the aluminum. And, when we’re all done, Tim’s going to have a great frame!

I’ve also nearly finished a fork for an old project bike. Actually, the fork could go to paint right now. But it has a stainless crown and tips – and polishing takes a long time. I was pleased to find out that other builders have the same challenge as me. This is, to see a remaining scratch from one level of grit (be it sand paper or buffing compound) requires polishing down a couple of more levels. So it’s a two step forward and one step back kind of process. I find that it’s important to be in the right frame of mind.

When working the stainless, pressure has an equal impact to grit. It’s possible to work a piece for a while, and with one stroke of two much pressure, scratch a layer that needs to be sanded out before proceeding down the levels of grit. This plays out especially with oddly shaped pieces (can you say lugs, crowns, and dropouts?) where an even pressure needs to be applied around varying contours. On a nice flat surface, I can work through a series of grits with a flap wheel sander and have few problems. The flap wheel sander works fast, which is nice. But through in a corner or a curve and it may be possible to sand through the piece before getting it uniformly ready to work down to the next level of grit.

So, I’m about halfway, maybe a little more, through polishing this fork. But it takes a peaceful/patient frame of mind to get through this work. With what’s in front of me, the fork is being set to the side until I catch up a little. Anyway, when it’s done, another frame will go to the painter.

Finally, I’ve been filing (on my trip) some old Cinelli pressed lugs and have them almost ready to go. This is for a track frame I’m building for my neighbor. He want’s something very traditional. He’s got a Record Pista crankset, some nice pedals, a Cinelli steel stem (and soon a matching bar). He’s working on building up a set of wheels with some Record Pista high-flange hubs and some nice old Campy rims.

The Cinelli lugs are more symmetrical than some of the pressed lugs that I’ve used. But, in the weld area, near the edges, there is some porosity. We’re going with a cast Cinelli bottom bracket which should arrive this week. Hopefully it’s in a little better shape.

For tubing, I’ve been considering lots of options. The rider prefers to stick with Columbus – which is fine with me. However, there is a limited variety of standard-sized tubing available in the newer metals and these tend to have thin walls. As a consequence we’ll probably use one of the old sets of SL that I have kicking around. While I grew up thinking that SL was dangerously thin and whippy, it’s seriously a heavy gauge tube compared to modern tubing. If an SP downtube can be found, that will probably get subbed in. Also, we’ll look at the option of going to a heavier gauge chainstay. Together, this should provide the rider with the stiffness he desires while sticking to traditional tubing diameters.

My plate is full, eh? Also, I need to do some reconfiguration of the shop to make my space more efficient, and I’m starting to look at lathes. There’s lots to do so I’ll stop writing for now and get back to work. See you soon.

Tuesday, August 28, 2007

Fork & Crown

Fork & Crown. Could be the name of Pub, eh? Anyhow, I've had forks and crowns on my mind lately and there are a few pictures to share.

The last post contained a bit of blather about the difficulties of polishing stainless - enough that you might wonder why I or anyone else does it. Hopefully this picture answers the question. The crown on this fork is nearly done. Click on the pic to enlarge it. This was taken with an older 2 mega-pixel camera, without flash, this evening at my kitchen table. You can clearly see the camera and towel (on which the fork lies) in the finish of this crown. Just a little more touch up here, and a bit of work on the dropouts (also stainless) and this can go to the painter.

Here is a view of the complete fork. It's straight legged and the dropout sockets have a similar shape and cutout as the crown.












Along with polishing, this evening was spent filing the intersection of some seat stays and a seat cluster.
The cluster is rather unique (in my experience) in that it is fastback with legs coming off of the binder bolt tabs all as one casting. These are designed to fit a bi-conical stay, and start rather square, then shift towards round where they mate with the stays. Anyhow, the legs needed a little tuning to make a smooth transition into the stays. I think the overall effect will look very nice under paint.

This seat cluster above is on the Path Racer I'm making for myself. This frame is built with an old set of Reynolds 531 using Prugnat lugs. The fork crown is an old full-sloping casting, and was very rough when I started. The most
frustrating part of the crown was that the spigots (over which the fork legs slide) were very asymmetrical, with one side of each being thick and one very thin. And, the thin side of each was facing the same side of the crown. So, it was impossible to file them to fit the fork legs while maintaining an equal distance from the center of the steerer tube for each leg. The difference probably won't be visually apparent - and the fork with is well aligned and will steer nicely. Still I didn't like the result.

Just to quantify things, I'm close to having an old set of pressed steel lugs prepared for use - maybe 4-5 hours worth of work. But the old sloping crown has already taken at least twice as much effort and I'm still working on its final finish. The problem is that the fork legs are Imperial Ovals.

Richard Sachs makes a point that new builders, working with todays high quality castings, just don't appreciate what long-time builders have gone
through as part of their learning curve. He's right, until one works with old time castings and pressed parts, its hard to understand how much goes into preparing these, parts prior to building.

One also loses some of the opportunities to develop ones eye, and tool skills, that comes from doing significant fabrication work on these parts. This is one of the reason for my recent focus on old lugs and components. The modern stuff is undoubtedly better, but working with the old stuff is improving my skills.

Because I have another set of 531 w/ Imperial Oval blades, and I prefer a flat topped crown, I contacted Richard to see if he had any kicking around in his shop. He was gracious enough to send me one gratis - and its a Nervex to boot.

You can see that the old Nervex crown is considerably narrower in overall width than the newer Continental Oval crown (today, crowns and legs are made almost exclusively in the Continental Oval style). These two crowns have the same spacing between the fork legs, but the Continental crown is clearly wider. overall.



Many folks hear Nervex and think of the Nervex Professional lug set. No wonder, with its intricate design, it was a builders favorite for top end frames including the Schwinn Paramount. But Nervex made a variety of styles of lug, fork crowns, and bottom brackets. The crown that Richard sent me is not the Professional, although I'm not sure what model it is. Suffice it to say that it is an example of a quality crown of its time. Here's a picture.

What may not be clear are some file or grinder marks just to the left of the center bulge and just to the left of the right hand socket. At first I thought that Richard had sent me a crown that he had previously started to work. I related to him that it was difficult to start in on a crown that showed his handy-work. He let me know that this crown, complete with these marks, was as-delivered by Nervex. Perhaps this is where the mold connected to the sprue, and the channel for steel to enter the mold. But its interesting, especially given the rough nature (compared to today) of the casting, that someone had to perform some manual clean
up before the crown was shipped.

In the picture of two crowns, the second on is a Sachs Newvex in stainless. Because it will be polished, and because it has some complex shapes, it takes more preparation time then most newly manufactured crowns.
But, make no mistake, this is a very nicely made crown - any builder would be happy to work with materials of this quality!

As we look more closely, its easy to
see how much thinner the walls are on the newer crown. The old castings required significant thinning by the builder before they could be assembled into a bike. This is of course all done by hand with a file. And, in the process, the builder has to maintain fidelity to the shape of the crown. Curves must remain smooth and even, for example. And, in some places, its not possible to move the file around the crown on the desired path without hitting or rubbing another part of the crown. So a builder needs to employ creative approaches for getting his/her file into and around all the complex shapes of the crown.

Another difference between these crowns is that the old one has built-in reinforcing tangs on the inside of the blades, whereas the new one has
pockets into which separate tangs can be place prior to brazing. One advantage of the modern approach is that the tangs can be prepared with cutouts apart from the rest of the crown - which makes things easier.

Going back to the first picture of the pair of crowns, the Newvex crown has
an extended area that has been throttled in preparation for cutting the crown race. If the builder is using a crown race cutter (as opposed to a lathe), then this will need to be further reduced with a file prior cutting the race. Skipping this step forces the cutter through too much material, with the consequence that it wanders some from side to side. In the process, the crown becomes too narrow and the headset won't be held firmly.

On the Nervex crown, the throttled section must be continued down to an appropriate depth by the builder prior to filing and cutting for the headset race. Depending on a builders skill, this could take a considerable bit of
time.

Here's a side view. The Newvex crown is based on the old Nervex
Professional and you can see all the frilly detail for which the Professional is and was known. By comparison the Nervex crown is rather simple. I may yet shape it to place more emphasis on the outer two points. On the Newvex crown its easy to see the pocket (indent) for its add-on reinforcing tangs. Also, if you look at the bore for the steering tube, the Newvex crown has a ledge on the bottom providing a hard stop for the steerer prior to brazing. The Nervex crown is open on the bottom, which gives to builder the option to have it stick out a little as a platform for applying filler. Then when the brazing has cooled off, the builder can file off the excess steering to be flush with the crown itself. Finally, the Newvex crown has a breather hole. By putting a matching vent hole in the fork leg, the vent hole is: a) out of sight; b) surrounded by, and brazed to, the crown - providing more strength to this invasion of the fork leg. The Nervex crown has a huge opening that can act in the same way as a breather hole. Because of the limits of old-time casting, however, the fork leg gets much less support on this side than with the newer crown. Furthermore, because the bottom of this opening terminates in a sharp V shape, there is the possibility of inducing a stress riser. Is the crown or leg likely to fail due to this rise? Probably not. On the other hand, modern casting abilities prevent us from having to test this assertion.

Ah, once again its late. So I'm going to sign off here. Happy Trails!