Hardening and Tempering

Of all the basic processes used in the small workshop none is more useful than the beat treatment of metals and, I would venture to suggest, none is so imperfectly understood. Heat treatment of certain metals if correctly applied can modify their physical properties to such an extent that a piece of steel can be used to cut itself, or a copper wire can be drawn out so finely that nearly seven thousand yards are needed to make a one ounce weight.

The most common use of heat treatment by the clockmaker and model engineer concerns the hardening and tempering of steel. Both terms are in common use and are often confused with each other. It is for instance quite common for the layman (and, dare I say it, for the technician who should know better!) to refer to an object as being made of "tempered steel" as though this is an indication of superior quality. In fact we must be quite sure what we mean by this and other heat treatment terms right from the start. Most metals in their natural or raw state are fairly soft; this indicates that they may be bent or that a sharp point will make an indentation into their surface. By suitable heat or other treatment this softness may be modified so that the metal no longer bends so easily without fracture, or it resists indentation or scratching. It is then said to be hard and although the degree of hardness varies from metal to metal and with the actual heat treatment in many cases, steel for example, can become hard enough to cut glass. Quite often the hardness produced by a certain method of heat treatment will be greater than required causing brittleness for example. When this happens it is sometimes necessary to use a secondary heat treatment to remove some, but not all of the hardness. This process is known as "tempering".

As clockmakers virtually all our work will be performed on two metals only, viz: brass and steel, and it is with steel that we are most often concerned when we talk about hardening and tempering. Steel is a ferrous metal, ie one that contains a large proportion of iron. Iron itself is a relatively soft metal not capable of being hardened. However pure iron exhibits a remarkable affinity for carbon which it can take into itself in either liquid or solid solution in a variety of forms. The metallurgy of iron/carbon compounds is quite complex and has formed the subject of many complete volumes but for our purposes all we need to know is the effect that varying degrees of carbon have on some of the physical properties of iron and steel, in particular the hardness and the way in which this may be affected by varying heat treatments.

As it comes from the blast furnace molten iron contains a high proportion of carbon, usually in the order of about 6-7%. This is often partly in the form of free graphite and the molten iron solidifies to form the well-known "cast iron". Most of us are familiar with cast iron; it is fairly soft, quite brittle, and although strong in compression it dislikes bending or tensile stresses. However, by a variety of processes, the carbon contents of the rudimentary cast or "pig" iron is gradually reduced until we are left with a product containing the requisite amount of carbon. Quite often in fact the carbon content is reduced to well below that finally required and then brought back to the required level by very careful metallurgical control, using extremely sophisticated equipment.
With little or no carbon left in the iron we find that the metal becomes soft and malleable, ie it responds to blows from a hammer even when cold. Of course when brought to a bright red heat it becomes quite plastic under the hammer and this is the raw material of the blacksmith. Pure "wrought iron" with a very low carbon content is now only made in very small quantities, much to the annoyance of most blacksmiths who can achieve a much higher quality of work by using iron with a very low carbon content.

The material most commonly available is known as "mild steel". This usually has a carbon content of about 0.3% which gives it a slightly greater strength than wrought iron but makes it less ductile and malleable. This is the steel used to make structural sections such as rolled steel joists and in the cold rolled and drawn versions is the "bright mild steel" used so much in industry.
Steel with a carbon content of around 0.3% -0.7% is usually known as "medium carbon" steel. This is a much tougher material than mild steel and is used for such things as car axles and similar highly stressed parts. Specialised heat treatment is sometimes used on these steels to obtain maximum toughness, but this is not a process commonly to be found in the small workshop. Medium carbon steels can be useful to the clockmaker for such things as pinions and pivots, particularly in the larger sizes.
As the carbon contents of the iron approaches 1 %, it is referred to as a "high carbon" steel and a distinct change takes place in its characteristics. For the first time in the ascending order of carbon content we find we have a steel which can be hardened in the true sense of the word.

The hardening process is carried out by heating the metal to a high temperature, usually a bright red heat, and then rapidly quenching the red hot steel in water, oil or other suitable cooling agent. After this process it is found that the steel has become extremely hard right through and is in fact hard enough to make a cutting tool suitable for use in the lathe or other machine. The actual degree of hardness will depend on the composition of the steel and the rapidity with which it is quenched.

So much far the principles - now we need to see how the process is put into actual use in the workshop. Firstly we must decide on a suitable material. High carbon steels are readily available in industry but the clockmaker may experience some difficulty in finding suitable supplies. The most commonly available high carbon steel is known in this country as "silver steel", presumably due to its polished appearance since it contains no actual silver. On the other side of the Atlantic a similar material is known as "drill-rod" although this in fact is of a slightly superior specification. For many years silver steel was obtained in 13 inch or 6 ft long rods in a wide range of sections both round and square. Both fractional inch and Morse twist drill sizes were stock items and the material itself was usually sold by weight. Now in these more enlightened days we are only allowed to purchase a restricted range of metric round rods, priced by the length, all other sizes including square rods being "non preferred" and virtually obsolete.

Another method of buying high carbon steel is the form known as "ground flat stock". This is a steel prepared for use by tool and gauge makers and is often known as "gauge plate". It is sold in 18" lengths, in a wide variety of rectangular sections including squares, from'/64" x 1" right up to 1 '/2" x 6". It is much more expensive than silver steel and usually comes in packets containing comprehensive heat treatment instructions. Some of the larger steel suppliers such as Macreadys still stock a rather old-fashioned "water-hardening cast steel" used for making chisels etc and if this can be obtained it is often a very cheap way of finding suitable stock. And finally, it must not be forgotten that old files are made of first grade high carbon steel, and the material from these should always be saved for future use.

Apart from old files all other grades of high carbon steel will be in a soft or "annealed" condition when supplied. This means that the steel will be soft enough to be sawn, drilled and filed to bring it to the right shape for the tool or component we wish to make and, apart from the final polish all cutting and shaping should be completed before hardening is attempted. The hardening is carried out by bringing the steel to a bright red heat and then quenching it in the chosen liquid. Many older books describe the correct temperature as "cherry red" but this of course does rather depend on the cherry! The actual temperature required is of the order of 770° - 800° C which corresponds to a bright red heat when viewed in daylight but out of direct sunlight. For those who are unsure of even this temperature it is a useful but little used fact that at 769° C steel reaches what is known as the Curie point, at which it ceases to become magnetic; thus a strong magnet brought close to the steel while it is heating up will readily reveal when the critical temperature has been passed.

If the steel is heated much above a bright red heat there is a danger that some of the carbon will burn off as oxide. This of course will reduce the carbon content of the steel, possibly even reducing it to such a level that it will no longer harden. The old-time blacksmiths were well aware of this and referred to steel to which this had happened as being "burnt". Peter Stubs, the well known Lancashire file maker, used to guard against this by coating his files during heating with the residue found at the bottom of the beer barrels used in his other main activity, that of a brewer; the "barm" from the spent yeast carbonised and diffused an extra layer of carbon into the red hot surface of the metal, thereby producing the hardest files available at the time and making both his name and his fortune.

Another point to be given consideration is that the whole process takes time. Once the correct red hot temperature has been reached the steel MUST BE MAINTAINED AT THIS TEMPERATURE long enough for all the requisite physical changes to take place. The usual rule of thumb that is quoted is "one hour for every inch of thickness" although this may prove to be a little on the conservative side. In practice I find that a typical tool made from a piece of 1/4" diameter silver steel needs to be maintained at red heat for about one minute to ensure satisfactory hardening.
When the steel has been heated to the correct temperature for a sufficient length of time it must be quenched. In the past, before the metallurgy of carbon steels was completely understood, there were many myths and legends concerning the best medium for quenching hot steel. Sheep's urine figured largely in many of the recipes and, for the piece of steel that obstinately refused to harden properly, one writer solemnly assures us that it should be quenched in the urine of a redheaded boy! These days we know that poor hardening is much more likely to be the result of incorrect carbon content or a faulty heating procedure but the quenching does have a definite part to play in the hardening sequence. Cold clean water will be as satisfactory as other liquids for most purposes, although some authorities recommend a 10% brine solution for slightly quicker heat dissipation and hence greater hardness.

Rapid quenching does have some disadvantages. In articles with sharp internal corners it is possible for stress cracks to be set up which can, in some cases, be highly undesirable. Other features of rapid quenching which are noticeable particularly in the lower grades of carbon steel such as common silver steel are that both distortion and dimensional changes can occur. Neither is very desirable particularly if accurate tools or gauges are being made. For this reason most ground flat stock contains other special alloying elements such as chromium and vanadium to prevent cracking and distortion and the steel is designed to be hardened in oil rather than water. It is often sold under the label "lowdistortion oil-hardening steel" for this reason.

Many of the tools and components we wish to make for our clockmaking activities will be long and slender such as drills and pinion arbors and warping or distortion is to be avoided if at all possible. A useful tip here is to plunge the red hot component VERTICALLY into the quenching bath. An even more effective precaution is to hold the work in the drill chuck of a vertical drilling machine and bring it to red heat while it is rotating fairly slowly. Then when the correct temperature is reached the work can be lowered into the quenching bath while it is still rotating. This minimises any distortion and is particularly effective for long slender clock arbors.

Other quenching media are available of course. The most rapid quench of all is obtained by the use of mercury but, in view of the toxic nature of mercury vapour, the process is not to be recommended unless proper laboratory conditions are available. Another method much favoured by the old time watch makers for making tiny drills was to heat the drill in the flame of a candle and then quickly plunge it into the body of the candle. Speed was essential since the draught of air surrounding the candle flame could chill the work but not sufficiently quickly to quench and thus harden it.

One of the problems attendant upon heating iron or steel is that the metal oxidises and forms a black scale on the surface. In a lot of cases this is not particularly important, but where we are making small delicate clock and watch components which are difficult to polish subsequently, it is desirable to keep this scaling to a minimum. In industry the problem is overcome by heating the work in a reducing flame, or in some form of inert gas, so that oxidation cannot take place. On a smaller scale we can often achieve satisfactory results by taking steps to exclude air as much as possible while the work is being heated. One such method often used for pinion arbors is to bind the whole length of the work with soft iron wire, and then to coat the whole thing with common kitchen soap rubbed well in. When heated the soap melts and partially carbonises forming a thick protective coating which excludes air quite effectively; when the work is extracted from the quench and uncovered it will be found to have an even grey colour all over.

Another method often used for items like balance springs is to enclose the work in a specially made airtight box usually of thin copper packed inside with powdered charcoal. Air can be excluded by sealing round the lid or any openings with fireclay. The whole box is brought to the required temperature and then quenched in the normal manner.
After quenching the work should be extremely hard. To check this the easiest way is to try to use a file on it. Not, of course, a good file since if the work is truly hard it will be as hard as the file and will ruin the file teeth. The file should in fact skid over the work without making any sort of impression. If it is possible to get the file to bite at all then the work is not properly hardened and the heating and quenching sequence must be repeated.

For a very few purposes the work can be left in this ultra hard condition - such things as scrapers, glass cutters, china and glass drills need to be this hard, but steel left in this extra hard state is also very brittle and will snap or fracture at the first opportunity. To make the item of greater use we need to remove some of the extreme hardness and this we do by "tempering".
Tempering involves a second heat treatment but to a much lower temperature than that required for hardening. The temperature is this time much more critical, since the final degree of hardness left in the work will depend on the exact temperature reached. Fortunately, we do not need expensive pyrometers or other measuring equipment since we are able to make use of another of the physical properties of iron and steel, that of the colour of its oxides. As we already know when iron or steel is heated in air it oxidises on the surface. At relatively low temperatures, well below that at which thick scale starts to form, the oxides take on a distinct pattern of colours, starting with pale yellow, progressing through browns to purple and finally blue. And, as sheer luck would have it, these colours correspond exactly to the range of temperatures we require for tempering our steel.

And so to temper our piece of hardened steel we can adopt the following procedure. Firstly we polish it to remove any scale formed during the hardening process; this will enable us to watch the oxide colours forming. Next, we heat the work very gently in a suitable flame. What is suitable? Well, it all depends on the size of the work but it must be a clean flame. For much of the work I do on clocks I find a small methylated spirit lamp quite adequate, but for larger pieces such as big steel cutters one of the new pencil flame butane torches is very suitable. As the work is gently heated, a band of colours will be seen to form on the polished steel starting first with a very pale yellow, usually referred to as "straw" and progressing through browns and purple until a pale blue is reached, after which the colours cease to appear and the work becomes red hot again.

The secret of correct tempering is to arrest the action when exactly the right colour reaches the cutting edge or active portion of the work; this we do by quenching the work in water or oil just as we did when hardening, but here the quenching plays no part in the process it just stops the work from becoming any hotter. In fact, if a good oven is available or the hot plate of a solid fuel stove such as an Aga can be used it will be found that these constant heat sources can produce a far more even temper than an open flame.
For the record the actual temperatures required for common tempering colours are as follows:

Light yellow 227° C or 440° F
Dark Yellow 250° C or 480° F
Brown-purple 270° C or 520° F
Dark purple 288° C or 550° F
Dark blue 300° C or 570° F

The oven on the average domestic gas or electric cooker when turned up to the maximum can usually achieve 450° and sometimes 500° F and so this can be a useful method of tempering tools, although it is not usually hot enough to achieve a nice blue for colouring clock hands.

The actual degree of temper depends on the use of the finished item. In general lathe tools, drills and metal cutting tools need to be tempered to a light or medium straw colour; items which need more toughness and resistance to fracture such as screwdrivers and wrenches are best tempered to a dark brown or purple colour.
Occasionally work which has already been hardened needs to be rendered soft again so that it can be worked on. A typical case in clockmaking is where a pinion arbor needs to be drilled along its length so that a broken pivot can be replaced. Here the process of softening is referred to as "annealing" and it is carried out by heating the work to above the critical temperature ie to a bright red heat and allowing it to cool as slowly as possible. In the case of large sections this can often be achieved by allowing the work to remain in an open coal or coke fire overnight and retrieving the work from the ashes the following morning.

Finally we must mention the other common metal used in clocks, brass in its various forms. Brass is non-ferrous and cannot be hardened by heat treatment in the same way that carbon steel can. It will however harden by mechanical pressure and this is usually known as "workhardening". When brass is received in the soft state, such as in castings or soft sheet, it is often required to harden it before it is suitable for our work and this we do by hammering or rolling it.

Conversely there are occasions when hard brass needs to be softened, perhaps so that if can be bent or riveted, and here heat treatment is used to anneal the brass. The metal is heated to a dull red heat and allowed to cool and this treatment will be found to soften the metal completely. In some books it is suggested that the brass should be quenched in water after heating but this is purely a matter of convenience in handling; rapid quenching will have no effect on the annealing process.

This article, written by "Counterbore", was taken from the first number of "The Clockmaker" (April/May 1990), a magazine by Tee Publishing which unfortunately ceased to be printed