Saturday, 28 January 2012

Harden the four cup(s) *

I've been able to get out in the shed a little this week. Enough to fettle the rear hub from a bar of 4140. When I checked the drawings I had made back in 2008, I was certain that I'd made a mistake measuring the original and that I'd got the dimensions too small. I even checked a couple of 1890's front wheels I have lying around which confirmed that the measurements were correct. The bearings seemed tiny when compared with the large front wheel bearings, but are exactly comparable with a modern front wheel.

The first task is to make the hub shell, this uses the same techniques as in the front hub just on a smaller scale and with fewer bearing surfaces (!). There is no need to mill key ways to lock the gunmetal flanges in place since the rear hub transmits no torque. In practise I have found that the spoke tension will adequately hold the flanges in place, however I will either make them an interference fit or soft solder them on as was the norm back in the day. The oiler hole is tapped at 2BA, I'm happy using BA threads since they were first formulated in the 1880's and I have previously found them on antique bikes.

The next job is to make the cones, the fixed (right) cone is left smooth but the adjustable (left) cone is knurled with a straight, fine pattern. My knurling tool has diamond pattern wheels and with the shocking local price of replacement wheels, I've milled 80 'knurls' instead of rolling them.

Roughing out one of the cones

Parting it off.

Machining the outside face.

Milling the 'knurls'

The hub spins beautifully on it's 11 x 3/16" balls in either race. 

I've calculated a formula for the dimensions of the race based upon the number and size of balls required, this has worked out well so far although the formula leaves no room for a gap in the balls which they need. You can either add one to the number of balls (n) or just add a smidgeon to the calculated value D. I have found that making D larger by 0.020" or so is enough to get a useful gap. The calculated value for D in the rear hub = 0.1875/sin(180/11) + 0.1875 = 0.853". In practise I machined it to 0.875" and have a sufficient gap, as can be seen in the above photo, so the balls don't bind.

With all the steel components of the hubs now complete, I can get them heat treated. I initially chose 4140 steel because of it's suitability for nitriding which produces a very hard surface case up to Rc 64 but retains the tough centre. Nitriding will also cause far less distortion due to the lower heat required. SSM have a local office so I gave them a ring in the morning and was told that if I can get the parts to them by 10 o'clock they could put them in today's batch and I could pick them up in the afternoon. A fantastic service, I'm sure you'll agree.

The parts were still warm when I picked them up, 
like fresh hot cross buns straight out of the oven, only not as tasty. 
And a bit harder.

In other news my parents are now here and are pleased that I've been unable to organise any large aftershocks so far. My Dad is being kind to me and has only offered advice on using the dividing head to date. He sits on the stool in the corner of the workshop like a benevolent owl. I'm going to try and get the forks brazed up whilst he's here. He has far more experience than me at brazing, so I'll get him to do it and then I can examine his work. This will be the focus of my shed time next week.

We had a huge electrical storm at the weekend, It passed directly over our house and we suffered a very close lightening strike. Had I any hair remaining I'm certain it would have stood on end from the electrickery. My wife's new computer (birthday present) had it's motherboard fried and we are trying to arrange a repair on insurance. What fun. In the meantime I've resurrected an ancient steam powered computer to allow us to access the interweb and me to continue posting this nonsense.

* with apologies to Chopper Reid

Wednesday, 18 January 2012

The Front Hub

I've spent a little time this week making the front hub for the facile. This is a complicated affair and has 6 bearing surfaces that must all be concentric. It is hollow and has several holes and ramps that all need to accurately located. Lots of 4 jaw and fixed steady use.

Once again a series of photos to illustrate the various processes, click for larger versions.

Centre the ends, machine a plain surface to use the fixed
steady on and bore a pilot from either end. The pilot holes
met accurately (not too easy with a 7" bar)
Cut the first pair of bearing surfaces,
the end of the hub is also now at final size.
Now enlarge the bore to final size from this end.
Then machine the inner bearing race and clean up the bore,
These three bearing surface are now all concentric.
Now flip the hub around and repeat on the other end.
The centre section needs to be thinned to size.
Like this.
Then all that's left is to drill the holes to poke the balls through,
mill the key ways to lock the gunmetal flanges...
... and mill the angled ramp that is used to lock the sun gear to the  hub.
Finally after all that I can trial a fit up with all the parts assembled.
The hub spins beautifully on it's 60(!) balls.

In other news, my parents are arriving for an extended holiday with us. This has meant a flurry of activity around the house. The workshop has been tidied, the garden weeded, the lawn cut and the children polished. I only hope that the work so far on the geared facile bears Dad's scrutiny. I just know he'll be out fiddling with stuff whilst I'm at work. I'll have a word with Mum.

I reported last time that I didn't feel the magnitude 6.0 aftershock as I was riding home at the time. It appears that I did feel it, I just didn't realise it. On the way home I had caught up with a mate of mine, another local cyclist, at a set of traffic lights. Matt has a new fangled bike computer that does GPS stuff as well as speed etc.  I'm a retrogrouch and only need a calendar to calculate my average speed these days. This means that given the exact time of the 6.0 from geonet, he was able to work out very accurately where we were at the time. At the spot in question we were almost home and on a very wide straight road so we were riding two abreast and chatting. I remember that very suddenly we clipped handlebars for no reason, I dismissed it as either Matt, myself or both of us being a crap cyclist. Now Matt's a handy rider so that just left me. I'm just glad that my ego is intact.

Friday, 6 January 2012

Smells like Dads

Now then, where did we get to? Oh yes, I'd machined the bearings for the front hub. Unfortunately at this point we had another earthquake swarm, we took a holiday and my wife fell off her bike so I haven't updated you for a while. Near normality has now been resumed.

The outer bearing races need the castellations machining. These act as the locking points for fine tuning the bearing adjustment. There are 35 such castellations and combined with a 26 tpi thread means that advancing one notch adjusts the bearing clearance by 1/910th of an inch. The castellations are machined at a 45 degree angle so that the locking tab, also machined at the same angle will sit flush in it's locating groove. More later.

Holding each race to machine these notches proved tricky, I initially tried to grip the threaded area in a three jaw chuck, protecting the threads with a soft copper strip. This approach proved unsatisfactory on so many points that I quickly gave it up. The machining tended to pull the workpiece from the chuck and clearance was not possible where the jaws gripped. Also three jaw chucks aren't accurate enough to properly centre the workpiece for the tolerances I wanted. So I needed to made a one off chuck to hold the races. First I machined the thread to screw the chuck directly to the Myford's register, the Myford nose has a 1.125" X 12 tpi thread. Then with the new chuck mounted onto the nose I could machine the internal thread to hold the bearing race, this has the advantage that I am guaranteed that both threads are concentric and hence that the race will be held perfectly concentrically. I also need to bevel the outer edge to allow clearance for the slitting saw to cut the grooves.

This new chuck can then be mounted on the Myford dividing head (same thread and register) and the setup as per the following photo. 

Indexing the castellations with the dividing head mounted on a swivelling slide mounted on a fixed slide.

The cuts are made in one pass at full depth, the depth being advanced with the cross slide and the cut being fed with the fixed vertical slide. The swivelling vertical slide is locked in place with it's gib strips, in fact every slide not in use was locked to maximise rigidity on such a contrived setup. Also note that I needed to make an arbor for the slitting saw to use between centres. I did buy a cheap Chinese arbor that was so eccentric that it was unusable, the saw was only cutting on a few teeth at a time rather than all of them, this caused massive vibrations and chatter and much anxiety. Just goes to prove that you get what you pay for I guess. I had intended to put a grub screw and a copper disc at the bottom (same as myford does when locking a grub screw on a thread) into the body of the chuck to secure the races but I found this was unnecessary. I simply placed a rubber sheet on the bench and the race screwed into the chuck onto this, then applying downwards pressure whilst turning I could tighten or loosen the race from the chuck, the rubber gripping the race to allow it to be tightened securely. The actual machining took just a few minutes per race whilst making the chuck, the arbor and the setup took many hours. I'm pleased with the results though.

The next component to be made is the hub that rotates in these races. First though, I need to machine a test piece to ensure that I have the clearances correctly calculated. The design of the bearing is such that the balls are loaded by inserting them up through a hole into the middle of the bearing. The placement of this hole is critical and only has a 5 thou margin of error. Too close to the outer race will impinge on the bearing surface and too close to the centre will not give clearance for the balls to be inserted past the lower flange of the race.

When I was measuring the original machine, I wondered what this hole was for. It's only as I came to draw up the bearing design that I realised that it can't be assembled without it.

Here's one I made earlier.

The complete bearing is loaded as follows

Screw both races into the bearing housing, the inner race in it's correct location and the outer race fully screwed in up against the inner race.

Load 15 X 1/4" balls into the inner race, there's plenty of clearance at this point so these can all be loaded from outside.

Hold the inner race up against the balls so that they don't fall out and orient the hole to be at the top.

With the outer race screwed fully in against the inner race there is just clearance (the 5 thou mentioned earlier) to poke the remaining 15 balls into the outer race through this hole.

Then screw the outer race out to make contact with the balls and lock the race with the castellations. The inner race is locked with a grub screw.

The inner row of balls is protected by a labyrinth seal partly formed by this flange.

I'll make the actual hub next week.

In other news, our festive season here has been punctuated by a series of disasters.

1) The pre Christmas revelry was rudely interrupted by another swarm of nasty aftershocks, lots of magnitude 5 somethings and one magnitude 6. I was riding home when the 6.0 struck and I didn't feel it, weird eh? We have lots more cracks in our house and the workshop was thrown around a little. I was making a final cut one morning when we had a 5.1 or so and my lovely machined surface became a seismograph, I think I'll leave it as is. Things are beginning to quieten down a little now but I've still been woken three times this week, we've had several hundred aftershocks in the two weeks since Christmas.

2) Fortunately we had a holiday booked up in the Marlborough Sounds between Christmas and new year, we were going to walk part of the Queen Charlotte Track and stay at lodges rather than camp, The drive up was in Beautiful sunshine and as we started the walk it began to rain. It rained (hard) for the next 4 days non stop but then cleared up just as we were driving home again. Nice.

3) For Christmas I'd bought each of the kids a computer for their mountain bikes, so they were keen to try them out. We were enjoying the track at McLeans Island when my wife had a small spill. Very low speed and she'd managed to get a foot down so it didn't look too bad, turns out she needed 14 stitches. Of course, we were miles from a road and the ambulance had a bit of a nightmare getting to us but the St. John dudes were fantastic, very calm, very professional. Top marks guys. You begin to realise how small the population is in NZ when sat at the side of a mountain bike track. My wife seemed to know every other person that rode past, either that or she had taught them at some point.

6 inside and 8 outside

4) And worst of all, I ran out of Swarfega. I bought some more at lunchtime and had the tub sat on my desk at work. Tinkerbell, who sits near me, commented that "Swarfega smells like Dads". She is of course, absolutely correct.