A lecture presented before the Horological Society of New York
by Sigfrid Strommer
Mainspring manufacturing might well be classified as a specialty within the steel industry, therefore, my talk will be to a large extent about steel and its processing. To make a good mainspring is a long and exacting job. I am going to tell you about the extreme care exercised in the choice of materials and how they are worked, all the way from mine to mainsprings.
In our generation, watchmakers have always considered mainsprings as an ordinary staple article of commerce, easily obtainable at various low prices. It was not always that easy. Did you gentlemen ever stop to think where the early watchmakers of one, two or three centuries ago got their mainsprings, or how they made them, or of what they made them? Steel, as we know it, has been made only about 80 years.
Peter Henlein hammered out the mainspring for the first portable watch from a band of iron in 1504. That iron mainspring was crude, but it was the most valuable one this world had ever seen. for it was the beginning of the watch industry.
I have not been able to find any record of when mainsprings were first made of steel, but that must have been about the year 1550, when watches were first small enough to be carried in the pocket. Surely the .first wrist watch made for Elizabeth, queen of England, in 1571, must have had a mainspring made of steel.
Steel of some sort was known in Egypt 6000 years B.C., in China and Babylon 2000 B.C. Rather good steel was made in India in the year 340 B.C. At that time the Damascus swordmakers bought all their steel from Hyderabad, India. The record states that the ore was carefully selected, washed and roasted, if necessary. Then it was reduced with charcoal in small fireclay crucible and allowed to cool in the crucible. Next, it was subjected to a tempering process consisting of repeated heating while covered with an iron oxide paste to soften it to suit the purchaser, who regularly watched the process throughout. You may know that the test for a Damascus sword was bending it into a loop so the point touched the handle, then it had to go back straight by its own resilience. The demand for steel in those days was very small. It . was used mostly for swords, razors and surgical instruments.
About the 16th century, when watches were invented, the knowledge of steelmaking had spread to nearly all countries of Europe, but production and demand were still very small. Each maker developed his own methods and equipment. The steel was poor and far from uniform quality.
Quantity production of steel began in 1858 when Bessemer in England, and G. F. Goranson in Sweden, perfected the process for making open hearth steel and thereby started a new era of industrial development. Since then, rapid progress has been made and improvements in mainsprings have developed in a parallel line with improvements in steel.
It was about that time, in 1860, that Saunier published his Treatise on Modern Horology. According to his remarks good mainsprings were hard to find. He states that the metallic blade is very rarely homogeneous and worked with sufficient care to avoid different parts being of variable strength. Its energy alters with time and this change nearly always occurs irregularly throughout its length. The coils adhere and rub together, either permanently or occasionally.
The Bessemer steel of that date was plain carbon steel of a quality which would not be used now except for very cheap mainsprings. Since then, thousands of experiments have been made on steel. By adding special alloys entirely different qualities have been created. The methods for working the steel have also been vastly improved.
Let us consider how mainsprings are made today. I want to give you a broad outline of the processes, all the way from mine to mainsprings.
The manufacture of high grade steel has always been and still is based on raw materials of the highest quality obtainable, namely pure Swedish iron ore and charcoal for its reduction in blast furnaces. These ores in the raw state contain only a very small percentage of sulphur and phosphorus, two very undesirable ingredients of mainspring steel. The effort to reduce that content still further has led to the use of concentrates of ore in powdered form. A good deal of the constituents which have an injurious effect on the quality of iron is removed in the concentration. The powdered ore concentrate is carefully mixed in fixed proportions with a small quantity of charcoal slack. This mixture is then roasted into sinters, that is, spongelike lumps about the size of my hand. All this work of preparing the ore is· preliminary to the smelting into pig iron. In that process charcoal is used for fuel. I wish to state that only charcoal obtained from fir trees is used in the furnaces and no deciduous wood, charcoal from leaf trees, which contains considerably more phosphorus.
Steel upon which the severest demands are made is produced in Siemens-Martin open hearth furnaces or in the high frequency electric furnaces. In either case only charcoal pig iron of the highest grade, made in the careful way described before, can be used. Another factor of the utmost importance is that the men in charge of the furnaces have a wide experience in their work, because the production of quality steel is more of an art than a science, and the correct handling of the melt, from the time the furnace is charged until it is tapped, is more a practical than a theoretical matter. At the Sandvik Steel Works a melter is not placed in charge and entrusted with the responsibility of the steel making until he has worked about ten years at the furnace.
Whichever furnace is used, the ingots are always small, weighing from 400 to 1200 pounds. That small size has a favorable influence on the quality, because the steel in these small ingots is of a more homogeneous structure than in larger ingots.
The steel ingots are hot rolled into bars which in turn are hot rolled into flat strip steel about 1.16" thick. The hot rolled strip then goes to the cold rolling department. It is not safe to hot roll steel down to the thickness of mainsprings because of the danger of decarburization.
Cold rolled strip steel represents a highly finished steel manufacture. By means of cold rolling, the thickness can be reduced to an almost unbelievable degree, in fact about 3/100 of a millimeter, the thickness of tissue paper. Cold rolling also makes it possible to obtain very exact and uniform thicknesses. The mechanical properties of the steel are greatly improved by the rolling, and the hardness can be increased so as to produce remarkable resiliency in the strip. Another advantage is the smooth and bright surface of cold rolled strip steel. It must also be kept in mind that cold rolling is just as much a question of correct heat treatment as it is a mechanical process. The plant for annealing is therefore very important.
The heat treatment begins as soon as the hot rolled strip arrives at the cold rolling department. It is then subjected to preliminary annealing and pickling for removing the oxide skin which forms in the hot rolling. Then cold rolling commences and is continued until increasing hardness, due to the rolling, makes it necessary to anneal the strip once more. When the material is rolled very thin, as for mainspring steel, requiring a large number of passes, it might be necessary to anneal the strip up to eight times during this process.
All intermediate annealings must be carried out so that no oxide is formed on the steel. The strip should display as bright a surface after the annealing as before the treatment. The Sandvik Steel Works use furnaces of their own design for annealing in a protective gas, and also furnaces where the annealing takes place in hermetically, sealed chambers. Both types are heated electrically and equipped with automatic heat reglating devices of the most modern design, enabling the highest degree of precision in this heat treatment.
The Sandvik Steel Works started manufacturing hardened strip steel in 1883. They therefore have exceptionally long experience in this line of work. During the past ten years, practically all of their cold rolling mills, and hardening and tempering furnaces have been replaced with new ones of the most up-to-date design. All furnaces are electrically heated and equipped with automatic control instruments that maintain a constant temperature at any degree desired. They are equipped to manufacture hardened steel strip up to 42 inches wide and at the other extreme very narrow hardened strip steel with a minimum thickness of only 3/100 of a millimeter. The thinnest watch mainsprings used are still nearly twice that thick.
Mainspring steel is drawn through the hardening furnaces at a constant speed directly into the quenching bath. It is then cleaned and again heat treated, almost to spring temper. The final tempering is done after the strip has been cut into spring lengths. After tempering, the steel strip is thoroughly polished, well oiled and wound into coils of about 8 to 10 inches in diameter and weighing 10 to 20 pounds. The material is now ready for the mainspring department.
This strip is then cut to the exact length specified on orders. In the cutting process, the holes are punched at the same time in both the inner and outer ends. That requires very perfect dies made of special alloy steel. Take, for instance, a die for cutting mainsprings with a double brace and hole at the outer end. It must have five punches, one for the coil hole, two for the brace, one for the hole in the outer end and one for cutting off the strip. All of the punches must be exactly in line. Perfect guides are also necessary to carry the strip through the die in such a position that the holes will be exactly on the center line no matter how narrow the mainspring is. Very accurate adjustments necessary in setting up the dies in punch presses.
After cutting, the spring blades are carefully tested for hardness arid flexibility. The results of these tests are checked against tables or figures assembled from past experience and the exact degree for the oil-tempering is decided upon. After tempering we recheck again. Final results must agree with our standards.
The ends of the spring blades are next partly annealed and repolished. They are then measured with automatic gauges, accurate to 1/1000 of a millimeter, and are ready for reversing or for the crosscurving process. If intended for spiral quality mainsprings, they are practically finished. We need only form the coil at the inner end and wind them into the holders or rings.
When taken out of the boxes, they open up into very small, spirals. The inner coil end is now on the outside. The center coil is then partly formed and the springs are wound in the reverse direction. Consider the circumstances! In the first winding, the part which becomes the outer end was inside and was bent very sharply. The other end was not bent so much. In the reverse winding the inner coils are formed from the latter part and bent more sharply while the outer ends bend less. That equalizes the tension in all parts as nearly as it is possible to do so in flat mainsprings. The springs are then wound over several arbors to give the coils the desired form. Lastly, the end braces are riveted on and the springs are finished.
From cross curved mainsprings still more even power can be obtained. That is because thinner and longer mainsprings can be used, due to the 25 % extra power this improvement produces. In modern watches, the mainspring drives the barrel from five to as high as seven and a half revolutions. (In some watches of a special construction as many 'as 24 revolutions). From the great number of dynamometer tests I have made during the past six years, I find that if you can get seven or more revolutions with one winding, then the power is practically constant during the first 24 hours run. These tests have also proved conclusively that the cross curved form increases the power of mainspring, on the average of 25%. In very thin mainsprings the gain is greater, in thick pocket watch mainsprings slightly less. In thick and wide mainsprings, such as clock springs the increase again rises to 25 %.
Dynamometer tests also show the importance of proper lubrication of the mainspring. After it is inserted into the barrel and sufficient oil has been applied, that oil does not immediately spread over the whole surface of the mainspring. Put the barrel arbor in a pin vise or in a chuck in your lathe and wind it five of six times so the oil works into the whole surface of the mainspring and you can feel it acting smoothly.
Special research work is being done in U.S.A. and Switzerland on this question of proper lubrication of mainsprings. It seems that thicker greases than watch oil are more efficient. Some advocate mixing graphite and oil. This proposal has considerable merit but also dangers, as regards watches. It is still in the experimental stage.
The question always comes up, "Why do mainsprings break?" The range of reasons run all the way from metal fatigue to sunspots. I read some years ago about two young watchmakers, personal friends, who discussed this question. They decided to study it and make notes. Some 50 or 60 years later they found that both of them had had exceptionally heavy breakage every 18 or 20 years. These were years in which the side of the sun containing most spots was facing the earth and consequently we had many magnetic storms on the earth. Be that as it may, we do not have enough time to check that theory.
Quality mainsprings can not be made except from quality steel. If the slightest impurity remains in the steel, metal fatigue accumulates in that point until the mainspring breaks, because it is weakest and bends more at that point every time the mainspring is wound up. Rust spots have similar effects, though not to the same degree, because oil on the mainsprings retard their growth. But, let us consider a mainspring made as perfectly as possible. Why should it break? According to our accumulated experience, changes in temperature break more mainsprings than any other cause.
Do you realize how much action takes place in the steel when the mainspring is wound up? On the inside the material is compressed and spreads out, becomes wider. On the outside it is stretched and becomes narrower. I have here a piece of steel, ground flat, with square edges, it measures 12.85 x 2.40 millimeters. After bending it with a radius of 1/2 inch to a right angle, the inside measures 13.20. It spreads out .35 of a millimeter. The outside measures 12.40. It became narrower by .45 of a millimeter. The difference in width between the outside and inside is .80 of a millimeter. The same action, in proportion, takes place in a mainspring. Furthermore, in cold the mainspring contracts, in heat it expands and relaxes. When a mainspring is wound up it is under a great strain. If then there is a considerable rise in temperature and the steel must expand, some mole·cules are liable to separate and start a fracture.
Some watchmakers have a habit of pulling out a mainspring straight when cleaning it. That is a bad practice. It amounts to the same thing as if I were to straighten out this piece of steel I referred to before. The molecules in the material would have to take their old position again and you know that even a soft iron nail will break eventually if it is bent back and forth a sufficient number of times. The proper way to clean a mainspring is to dip a piece of tissue paper in oil, bend it around the mainspring and carry it through all the way to the center coil without bending the mainspring anymore than is necessary.
Dirty benzine is also a danger to mainsprings. It may contain some water. The benzine itself evaporates but the water remains in the pores of the steel. and causes rust.
In conclusion I will say that it no wonder that mainsprings break. It is very wonderful that they last so long. When you consider that this long thin strip of hardened steel is wound up tight around a small arbor in that little space in a watch barrel, and expands and is wound up again sometimes as many as 20,000 times before it breaks, and does not set, then I think it is a marvelous achievement to make steel so flexible and resilient. The mainspring in a watch is the most wonderful thing made of steel.