Staffs & Pivots

[david pierce]

The AKA TOM TMAC articles have been thought provoking. There is obviously a lot to be said about equipment, techniques and procedures in order to produce a part, which in this case is a finished pivot and staff. This got me thinking about what the finished product should be and what is the best way to look at what really counts. A simplistic view would be, if the part is perfect then there won’t be and problems. This is true, but there is also the reality that nothing is perfect. This unfortunate fact changes things to, what needs to be done in order to make the staff and pivot work properly. In order for me to see this I have to break the part into separate areas of thought and look at each one as a separate challenge. For me these are material, dimension, concentricity and finish. Out of these four categories material and finish seem to have received the most attention. Dimension ranks third and concentricity is rarely if ever mentioned. Here is my current thought process on the matter.
1). MATERIAL: This would include the selection, size before machining and hardness before and after machining. After hundreds of years of trial and error tool steels seem to be the material of choice. I am sure that over time almost every possible type of metal, as well as other materials, have been tried and ruled out as suitable for this application. As far as which tool steel is the best I don’t think matters. It must provide sufficient strength, hardness and mach inability to do the job. Tungston carbide has more stability than steel but has less tensile strength. It would be almost impossible to turn a pivot down to the required size without breakage. It would also be more prone to breakage from shock.
2). DIMENSION: There seems to be a lot of leeway in this area. This is good news for the person making the part. If ball bearings were manufactured to the tolerance of a watch staff they wouldn’t work. If the crankshaft of a car engine was checked by sticking it into the main bearing and it flopped over 5 degrees, the engine would knock so badly that it wouldn’t run. Most pivots are around .004 inches or .1mm. This dimension probably has more to do with surface area frictional issues than fitting the pivot into the jewel. If a pivot is .1mm in diameter and .5mm long, the surface area of the pivot will be .157 sq mm. If the pivot is .2mm with the same length, the surface area will be .314 sq mm. doubling the pivot diameter will double the surface area of the pivot, and create more frictional surface area between the oil and the pivot. Even if the pivot still fit into the jewel hole and was able to turn, this difference would affect the forces in the train and the timing.
With the exception of the diameter of the balance wheel journal, the other dimensions such as the taper to hold the hairspring collar, the outside diameter of the staff, the length etc., are not super-ultra precision tolerances.
3). CONCENTRICITY: This should probably be concentricity and roundness. Concentricity and roundness are extremely critical issues and are probably the least thought of. To gain a perspective on this thought if a pivot is out of concentricity by only .001 inches, it is out by 25% of the shaft (pivot) diameter. This will create undesirable wobble that multiplies as the diameter increases. This will show up in the balance wheel. This has a terrible effect on the moment of inertia of the balance. To a certain extent bad concentricity can be corrected by poising but this is at best a band aid. Due to the extremely small diameter of a pivot it is almost impossible (for me anyway) to turn something down to .1mm or .004 inches strictly with a single point cutter. Provided that a turned piece is not removed from a collet or turned between cup centers and is turned with a single point cutter, the pivot should be concentric to the balance wheel journal and the tapered hairspring collet journal. The problem for me is once I get to around .010 inches the pivot becomes too weak to support the cutter pressure and snaps off. At this point additional equipment and techniques are required to support the pivot to get it down to the final size. This then becomes a trade off because the additional steps using the lathe accessories and a Jacot tool will affect the roundness and concentricity of the part. I am not saying that this cannot be done by someone more gifted than myself but I have never seen it done. Everything I have studied on this through books, videos and practice ends up using additional techniques and equipment after the single point turning is completed.
4). FINISH: A lot of discussion and thought goes into polishing pivots but I cannot recall anybody stating why they spent so much time doing this. If the pivot is placed into a jewel hole and flops over 5 degrees, a polished pivot will flop over just the same as an unpolished pivot. My thought on this comes from looking at a pivot under a microscope. When a pivot is magnified it looks like a mountain range. Since the pivot rides in oil it is essentially a hydrostatic bearing and if the surface is too rough (unpolished) the oil will provide dynamic resistance against the shaft. Since the oil surrounding the pivot is put into motion by the surface irregularities of the pivot, the consistency of the balance wheel motion can be affected which can result in timing problems.
David

[Bob Tascione]

Hi David,
You address some important and valid theoretical concerns here. In practice though these conditions are not usually difficult issues to deal with. I’ll touch a little on the importance of concentricity but would first like to make a couple of corrections.
When you say:

doubling the pivot diameter will double the surface area of the pivot

I was under the impression that the surface area change in a cylinder is non-linear to change in diameter or length and that doubling diameter would result in an increase of more than double (edit: after scratching my head and thinking about it…like 3 times) the original surface area while doubling the length will increase surface area by less than double original surface area. This further supports a designers reasoning for using the smallest pivot possible.

Also normally no need for a taper on the hairspring collet seat.

Concentricity is extremely important but is not so difficult to achieve. When you say,

To gain a perspective on this thought if a pivot is out of concentricity by only .001 inches, it is out by 25% of the shaft (pivot) diameter

it does seem like a huge amount, and it is. In practice though no watchmaker that knows what he or she is doing would ever use a pivot that’s out this much. Under high magnification this eccentricity might stand out about the same as turning a .500 pivot out of a piece of bar stock with it being off center 25% or .125. Very noticeable!
It is important to turn the pivot down as close to the finish size as possible leaving a little for final polishing. When grinding and polishing it down to finish size in a two or three step turning process (changing setup) one will see any eccentricity long before reaching .001 under reasonable magnification.

It does take some hands on turning practice to achieve close to perfection work but is not so difficult as one might think to reach a point where parts are being turned out that meet the design requirements of the average lever escapement watch or platform escapement.

Adios for now,
Bob

[bernie weishapl]

Great post with lots of info. It does take a little practice especially to know when you have left enough to do the final polish. I know I did a lot of clock pivots till I thought I had it down before I tried anything on a watch. I know I did all my turning by hand before I would let myself use a compound slide to make parts and I still tweak by hand after I finish with the slide.

[david pierce]

Bob & Bernie,
First of all, thank you for reading this and putting some thought into it. I welcome all comments as this is how I learn. My line of thought on the surface area of a pivot is as follows: Surface Area = Pi x Pivot Diameter x Pivot Length. If I substitute dimensions for the pivot diameter and pivot length the result is as follows:
Pi = 3.14159, Pivot Lingth = 2mm, Pivot Length = 3mm: 3.14159 x 2 x 3 = 18.84954 sq mm
Doubling the diameter will give us: 3.14159 x 4 x 3 = 37.69908 sq mm
37.69908 / 18.84954 = 2
This means that going from a 2mm diameter to a 4mm diameter yields twice the surface area. If I missed something, please let me know. The dimensions I picked for this were picked for the convience of calculation and are not actual pivot dimensions.
The concentricity of a turned part is to a great extent determined by the headstock runout of the lathe as well as the runout in the collets. If the lathe and the collets are running true, or the part is turned on dead centers, the finished part won’t come out looking and functioning like a crankshaft. Because of the equipment and machining techniques available to people who make these parts, concentricity issues are generaly controlled. I see this as a matter of balancing the manufacturing process.
david

[chris mabbott]

Great info David and thank you for taking the time to publish it, very much appreciated

The problem I’m having is determining, from the vast array of materials available, is the type of tool steel to use, and the methods of hardening, before, after, air, oil, media cooled etc etc, I’m like HUH 😆

Perplxr has a lot of great video tutorials on the tube, one I’ve recently watched was a watch staff creation, and like in Bob’s tutorial, he did it mostly by line of sight, you lucky guys 😆

Personally, I find it easier to make a little diagram before as all the size info I need is at hand, because dimensions, numbers etc, simply float out of my head 🙄

[david pierce]

Chris,
If nothing else the post wil inspire some thought. There is so little that is discussed about this issue and correct or incorrect I felt that it needed to be out there.
david

[Bob Tascione]

David,
I agree…not much material out there on this subject. Good topic.
Also…
Thanks, I stand corrected! Your formula is correct. My formula working with radius (2pi radius squared x 2pi radius x height squared = surface area) includes the area of the cylinder ends which wouldn’t come into play here. Your formula excludes the ends which is what we want. I probably should have had a couple more cups of coffee this morning when I posted that! lol

My point above was to address the 25% concentricity issue for those new comers that may not realize these numbers are hypothetical for the benefit of discussion and to avoid discouraging those reading this who may be thinking about making their first balance staff by showing this amount to be an extreme which can easily be seen under magnification and completely avoided. If the headstock or collets have this much run out then it might be a good time to repair or look for another lathe. Turning between centers will certainly take care of a lathe run out problem though. Point being, turning an acceptably concentric balance staff on a decent watchmakers lathe isn’t that difficult to learn and actually becomes much easier after successfully making the first staff as many members up here can attest to. One can see whether a staff has unacceptable run out under magnification during the machining, grinding and polishing process and always choose not to use the part.

To Chris:

The problem I’m having is determining, from the vast array of materials available, is the type of tool steel to use, and the methods of hardening, before, after, air, oil, media cooled etc etc, I’m like HUH

If you harden and temper the W1 you have (in water) you will be able to turn it down to a smaller diameter than when in the unhardened state. You’ll need to make sure your graver is very sharp when dealing with this harder pivot. Softer metal can cause problem when trying to turn down to a few thousands of an inch. Also by heat treating and tempering first you will avoid the need to do this after fabricating the staff. There are some advantages at times to turning before hardening though but I think you may find it easier starting with blued steel. Good to practice on a piece of W1 first though. If you run out I’ll send you more!

Bob

[arutha]

Chris,
if you have trouble cutting blued steel then try a carbide graver, it is well worth having one anyway as it can get you out of trouble if you burnish a bit of pivot steel with a blunt graver (and you most likely will when you first try this).

Concentricity – I am no expert at turning balance staffs, I have now cut a few good ones (I still manage to produce a lot of scrap staffs ) and when I first started I read that you could start in a collet but to keep concentricity you should, when turning the staff around to cut the other end, fit it into a wax chuck to ensure good concentricity. Daryn showed me that it is possible just to put it back in the collet and not have to mess about with wax chucks.
This will only work if you know you have a good tailstock which lines up perfectly with the headstock. You part off a little more material than you would need and turn a point on each end. You then pull the tailstock up to it with a pointed runner in there and looking through a glass with the headstock spinning you will soon see if the material is concentric. Once you have completed the first half you can then turn the material over and do the same again. Again this will only work if your tailstock lines up well with the headstock, the headstock collet socket is spotlessly clean and your collets are spotlessly clean and in good condition.
If a beginner like me can turn out a good staff then any of you guys could. Be prepared to over-cut something and have to start again and be very careful with the pivots. Another trick I was taught was to use an arkansas stone to get the pivots down to final (pre polishing) diameter.
Hope this helps a little and stops you worrying about things so much
Paul.

[david pierce]

Paul,
The concentricity can be checked this way but I find it easier to do with a centering scope. First I put a dial indicator in the headstock chuck of the lathe and sweep around the outside shaft of the tailstock spindle. While this step was not necessary to check the part concentricity with a centering scope I still did it to see how far out of center the tailstock is from the headstock spindle. Then I mounted the centering scope into a chuck in the tailstock to look at the part as it was spinning slowly in the lathe headstock. If the tailstock is not on center the wobble can still be detected as long as the part is spinning (slowly) and the centering scope is in a fixed position. The only difference is the wobble will not show up in the center of the scope. If the scope is dead on center or not it will easily show any runout down to at least .0001 inches. One of the Youtube videos shows a similar technique lining up a hole location for a pilar plate. The pilar plate is fixtured in a lathe faceplate and gently tapped into the proper position. I think the Youtube video was made by the University of Oklahoma Horology School.
david

[arutha]

David,
thanks for that info, was quite interesting but unfortunately I don’t have a centering microscope
When my eyesight gets worse (which I have noticed it is doing ) and I start producing wonky balance staffs I will definitely consider buying one!
Paul

[Author]

Posts