Making Keys for Uilleann Pipes, part I
by D.M. Quinn
Making keys for the chanters and regulators of uilleann pipes is a hurdle that every maker has to get over one way or another. In this article, I propose to examine some of the commonly used approaches to making keys, and as far as possible to show examples of the keys themselves and of the techniques used in manufacturing them. I’m afraid that once again I’ve chosen a topic that must be dealt with in installments, so I hope that interested readers will bear with us over the next few months.
Requirements: what does a key need to do?
All of the keys on a set of uilleann pipes are “closed” keys, i.e. they are arranged as levers to cover and uncover their respective tone holes, and until some force (the player) acts upon a key, it remains closed. In most cases each key will have its own spring arranged to hold the key closed with sufficient force that the air pressure maintained in the bag while the instrument is being played will not push the key open. The pressure or force required to open the key must be somewhat greater than the force exerted on the key by playing pressure. The pressure required to hold the key shut must be great enough that no air leaks past it, although other aspects of key design have something to do with this as well. There is almost always some sort of gasket or pad which is pressed by the key against the rim of the tone hole to make an air-tight seal when the key is closed. (Readers may wish to refer to my article on springs for uilleann pipe keys: Vol. XXI, No. 2, Spring 2002.)
It must be possible to open the key when it is required for musical purposes. This means that the part of the lever which forms the touch of the key must be in a position where it can be comfortably reached by the player. The motion of the key must be sufficient to allow enough air to escape from the tone hole it covers to produce the desired note without “squeaking” or having the note jump into an undesired vibrational mode. It must be possible for the key to open and close very quickly. The conventional layout of regulator keys also means that it is necessary to arrange the placement and motion of the several keys so that they do not interfere with each other. That is to say, the act of opening one key should not cause another key to open as well. (We will be examining an interesting exception to this rule a little later.)
Whatever the keys are made of, or however they are shaped and mounted, they must be able to hold up under the rigors of normal use. Keys which are easily bent soon fail in use, and regulator keys especially need to be robust. A key needs to be strong enough at the point of attachment of its spring so that no bending or distortion can occur there due to forces associated with the spring. Keys which make noise, such as when the metal of one key hits another, soon become annoying and distracting, although I believe we have all heard captivating performances which have also included some clinking and clanking.
The motion of a key must also be arranged so that in its extreme open position the key does not place undue stress on the block it is mounted in. If the “stop” of a regulator key is within the block itself, it is all too easy for the force of an exuberant opening of the key to break off the top of its mounting block.
Traditions and Conventions: what has been done before?
There are essentially two schools of key making. By far the most common and popular arrangement is a key which fits into a slot cut in the wood of the chanter or regulator, generally in a “block” which is formed by a ring or annular knob left when the chanter or regulator is turned on the lathe. Such a key requires an axle pin, which passes through the mounting blocks and the key, determines its point of rotation and holds it in place in the block. Good instruments have been made with a tremendous variety of block shapes and sizes, and the way the blocks are formed becomes an element of expression for the pipe maker, as do the shapes of the keys themselves. Keys made to work in slotted blocks have most often been made of metal, although wooden and other non-metallic keys have also been made to work. Non-metallic keys are very rare, and probably for good reasons. In this article, we will only be considering keys made of metal.
The other school uses keys made from sheet metal. These are typically quite a bit broader than analogous “slot” keys. Often such keys are formed as channels, with part of the sheet metal bent to form tabs through which their pivot or axle pins pass. These tabs may be held between other bits of metal, which is the common arrangement for Taylor-style regulator keys, or simply fit over the wood of the chanter or regulator, the common arrangement for Taylor-style chanter keys. This approach has been used by many makers for regulator keys, as well.
In the history of the uilleann pipes the most makers have chosen to use slot-mounted keys, and there is naturally a far greater variety of shapes and designs of this type. I believe their greater popularity may be in part because slot-mounted keys were (and still are) the norm for most other types of woodwind instruments. Modern clarinets, oboes, saxophones, bassoons and so on almost always employ keys mounted between metal posts. These posts may be screwed into the body of the instrument, attached to it with solder, or fixed to plates which are in turn fastened to the instrument with screws or solder. This has been done successfully with uilleann pipes, as well. In almost all cases post mounted keys show the heritage of slot mounted keys, and tend to be similar to them in configuration and geometry. Still, the majority of the uilleann pipes made within the last 250 years or so have employed keys mounted in slots cut in the wood of the instrument.
Some of the earliest woodwind keys that may be identified as of the slot type were made by folding sheet metal. This manufacturing approach offers a few advantages, but also has some serious limitations. When sheet metal is folded it can take on a rigidity that can be used to advantage, but the practical limitations of how much shaping can be done with or to a key of this type seem to have worked against it, and it is perhaps the least frequently encountered method of key making.
The set from which this key was taken was made in the early years of the nineteenth century. This form is similar to the short keys sometimes found on Baroque-era oboes. In this case, two tabs have been bent to form the axle bearing, but the “planes” of the key retain the sense of the flatness of the sheet from which it was presumably cut.
Folding can give a considerable increase in rigidity, as compared to a key which is essentially flat. A few pipe makers have adopted this approach to making keys, with varying degrees of success. Some of the nicest keys of this type that I have seen were made by the Waterford pipe maker John Henebry, who was active in the early years of the twentieth century. In the photograph below we can see that the shanks of the keys were formed by the folding of the metal into two layers, and that the touches have been formed into spoon like shapes.
The Taylor brothers also used slot-mounted keys on a good number of chanters of a configuration which I refer to as the “long block” style, in contrast to the style more commonly associated with their name. On one such example, I found keys formed in exactly the same way as the ones found on the Henebry sets. There are a number of these Taylor long-block chanters in circulation, but all of the others which I have examined have had keys forged from the solid.
The techniques: 1) Forging
Forging is the process of applying force to a piece of metal to upset it, generally by hitting it with a hammer, to alter its shape in a controlled way. In skilled hands, the metal can be shaped with a truly amazing range of possibilities, but essentially what has to happen is that one end of a piece of metal is flattened out to form the pad or head of the key, while the other is spread out, and sometimes curved, to form a comfortable “touch.”
Historically, most woodwind keys are of non-ferrous metals, such as brass and a family of copper-nickel alloys known popularly as “nickel silver,” (although there is no silver in it, or “German silver,” although apparently it was first concocted in China), silver or even gold. Various bronze alloys have been used, but these do not seem to be very common. Iron and steel have also been used, but anything ferrous except certain alloys of stainless steel will tend to rust, and this is not a desirable trait in keys. Aluminum has also been used, but as far as I know, it is not generally worked by forging.
The approach to forging will vary slightly according to what metal is being used, since various alloys will have various degrees of malleability. Some metals are most appropriately worked when cold, others when hot. The brasses and copper-nickel alloys, the most commonly found metals for woodwind keys, are conventionally worked cold. When a piece of brass or nickel silver is upset, such as by bending it or by striking it with a hammer, it becomes “work hardened.” The effect is cumulative, and the more the piece is upset, the harder it will become, until it reaches a point at which the structure of the metal fails, generally by splitting or crumbling. When the metal becomes hard from being worked, it can be softened by a process known as annealing, after which it may be worked until it again begins to harden up, and the process is repeated as many times as necessary to bring the piece to its intended size and shape. Knowing how much upsetting can be given to a piece of metal before it needs to be annealed is something that is acquired only by experience. If too much working is done between annealing’s, the metal can fracture and crumble, becoming useless for its purpose.
The process of annealing relaxes the stresses that are set up in the metal by bashing. To anneal brass or nickel silver, the metal is heated until it begins to glow and then is allowed to cool. Some workers advocate quenching the hot metal in water, and some claim that quenching makes the metal softer than it would be without quenching. I am not an advocate of quenching, but other workers’ experience may guide them. I believe it does little or no harm to brass or NS to quench it after it has “gone to black,” i.e. after it has cooled enough to stop glowing, but I prefer to avoid quenching altogether. Ferrous metals are annealed in a similar way but must be allowed to cool very slowly for the full softening effect. Full softening of ferrous metals may require that the “glowing” temperature be held for a specific period of time, depending upon the alloy.
The tools required for forging are simple. At a minimum, one needs an anvil, a hammer, and some means of safely heating the workpieces for annealing.
Any hard flat surface can be made to serve as an anvil, but to paraphrase the late and much-lamented Mr. Natural, one should use the right tool for the job. A proper anvil is a fairly substantial piece of steel with a smooth surface capable of standing up under the rigors of being bashed thousands of times. The heavier it is the better. For the first twenty or so years of my career, I used a short section of retired railroad track as my anvil, the top surface of which I had filed and polished smooth. Some six or eight years ago I acquired a proper blacksmith’s anvil, an antique which I was able to buy at a fairly low price. Honestly, I was not prepared for the difference a proper anvil made. The great mass of the anvil meant that much less of the energy I was pouring into the hammer blows was being dissipated by the movement and resonance of the anvil itself. It seemed wonderful to me at first to see how much more quickly and easily the keys took shape on a considerably more massive anvil.
The other side of the forging sandwich is of course the hammer. I have always used a ball-peen hammer for the job. The standard ball-peen hammer has a head with one end flat (actually very slightly domed) and the other end more or less spherical. I keep two such hammers for forging: one with an eight-ounce head, the other with a four-ounce head, both with polished striking surfaces.
An important consideration, for both hammer and anvil, is that the surfaces with which the workpiece comes into contact should be smooth. Any nicks or cuts in the surface of either hammer or anvil may be mirrored or “coined” into the surface of the workpiece. Making the face of an anvil smooth can be a time-consuming job, but the actual surface area which in practice is required for forging is quite small, and one can get by with preparing only a small patch of the anvil surface. When I first acquired my blacksmith’s anvil, its top surface was severely damaged (one reason I paid as little for it as I did), and through many years of heavy use it had developed a depression in the top flat surface, right over the center of gravity of the anvil. I decided to skim off the top of the anvil in order to resurface it. Using a fly cutter on a milling machine, I brought down the surface about 3mm, that is, until all but the deepest nicks and gashes had been eliminated. The central depression was deeper yet, so I decided to leave it and use it. In the years I have been using this anvil, I must confess I have added a few dings of my own to its re-worked surface, but I still have a few square inches of nicely polished surface upon which to do my forging. What is left of the depression has also been polished, and I use the concave area to help with the doming of key touches.
The tricks of the trade of forging are many and varied, and it is probably not appropriate to go into them very deeply here. In any case, I am not sure it would be possible or practical to describe the subtleties of the process. It is perhaps enough to say that the way one hits the workpiece determines how the metal is going to be upset. The behavior of the metal will be somewhat different if it is hit squarely with the flat end of the hammer or in a glancing way with the ball end. The way one hits the metal is determined by what one wants the metal to do. Unfortunately, I don’t know of any way to acquire a really meaningful understanding of the forging process without doing a great deal of it.
The process of forging only brings the metal to a shape that is approximately that of a key, and there is almost always a good deal of hand work or further machining left to be done with the forgings (the forged workpieces) after the hammering is done. The finish of the metal will have to be dealt with, as repeated annealing and hammering will leave the surface of the metal heavily oxidized and bearing the marks of hammer and anvil. The rough forgings must be filed, milled, sanded or ground to shape, and then worked with successively finer grits of abrasives until the desired finish is achieved. The same or at least a very similar series of steps is necessary for all methods of key fabrication, and we will examine the finishing processes in some detail later on.
When a slot-type key is mounted on a chanter or regulator, it must be prepared with at least a small section of the key having parallel sides so as to fit neatly into the slot in which it will operate. The thickness of the finished key in this parallel section needs to be slightly less than the width of the slot, and this is one of the variables that must be considered in achieving a proper fit which will allow the key to work smoothly and dependably. Perhaps the easiest way to deal with this is to use metal that has already been brought to an appropriate thickness, and to avoid forging or upsetting the metal in the area that will form the bearing, i.e. the areas that will eventually fit in the slot. In my own practice of forging chanter keys, I use commercially drawn or rolled rectangular brass bar stock sold as a nominal 1/8” x ¼”, and for nickel silver I saw out similarly dimensioned strips from 1/8” thick plate. With this approach, I rely upon the accuracy of the thickness of the metal as done by the manufacturer. If the key is forged from metal of a different size or section, such as from round rod for instance, or if the key shape is sawn out of metal plate, some time and effort must be spent on establishing the parallel sides necessary for a proper fit in the slot.
The brass I use is CDA360, an alloy known as “free machining” brass, which contains a small amount of lead. This is, admittedly, not the very best alloy for forging, but with attention to the process, it certainly can be made to yield excellent keys. Other alloys, especially those which contain no lead, may be more malleable and allow for more upsetting before re-annealing becomes necessary. In my case the choice of brass alloy was made on the basis of the format in which the metal was available. I consider the ready adaptability of the 1/8” x ¼” rectangular bar format to use as a slot mounted key to be more important than the extra attention this alloy requires in working it.
It may be of interest to mention here a technique known as “coining.” This technique, as far as I know, has never been applied to the making of uilleann pipe keys, but it has been used in the manufacture of other woodwind keys, and it certainly could be applied to pipe making if a maker considered the necessary investment in tooling a worthwhile thing. Coining is the forcing of malleable metal into a die of an appropriate shape. This is the method traditionally used in the minting of coins, and it is clear that a fine degree of precision would be possible. It would require a commitment to a specific size and shape for each key, and making up a coining die for each key would represent a very substantial outlay of time and effort, which apparently has never been justifiable to makers of uilleann pipes.
My own approach to forged chanter keys incorporates a good bit of machining and soldering, as well as proper forging. I prefer to have the head of the key round, so it will fit neatly into a pad seat that has been milled with a round cutter. It is certainly possible to forge a key with a round head, but in that case a good bit of shaping with files or other means will be necessary. I prefer to forge the touch end of a key, and to hard solder a disc to the other end to form the pad.
The following series of photographs will show the process of forging chanter keys, as I do it.
I start with a piece of brass cut from 1/8” x ¼” bar stock, a few millimeters short of the desired finished length.
The bar, or at least one end of it, must be annealed before the forging can begin. I keep a tray of pumice stones to serve as a hearth, a safe place to apply heat, which I usually do with an acetylene and air torch. (A propane torch will do as well.) When it is time to make keys, I almost always work them up in batches. I like to anneal keys in a darkened room so that I can more easily see when the metal begins to glow with the applied heat. I play the flame over the batch of keys and withdraw the keys one at a time when I see them glowing sufficiently. The metal needs to be heated until it begins to glow. Brass will go through a series of glowing colors as its temperature goes up, but for our purposes the luminescence of the metal is only an indication that it has been heated to the point where its internal stresses have been relieved. I believe the necessary softening takes place at a heat just slightly below what is required to make the metal glow, but it is generally safe to assume that if the metal glows, it has been heated enough to anneal it. There is no need to heat the metal any more than is necessary to make it glow in the area where the upsetting is to be done.
Heating the metal far beyond the first glowing red may lead to structural deterioration of the piece. You do not want the metal to melt; it only needs to be hot enough to begin to glow. As soon as I see a key blank glowing, I pull it from the pile and set it aside to cool.
Keys with curved touches need to have a dog-leg put into them before the first pounding takes place. Some care must be exercised here, because it is possible to create a point of extreme stress depending upon the way the metal is bent, and a mark made by pliers in the surface of the metal may become the starting point of a fracture.
A similar consideration must be given to the quality of the surfaces that will come into contact with the hammer and anvil. If one is starting from rolled or drawn bar stock, the surfaces will be quite smooth and ready to begin bashing. If the piece is sawn from plate stock, it is advisable to spend a little time smoothing the surfaces to remove the roughness and irregularities of the sawn surface. If a rough surface is forged, it will very quickly appear to become smooth, but small cracks may also form, and it will take longer to finish by filing and abrasives than a similar piece which started out smooth before being hit.
After the workpiece has been annealed, it is at its softest state. When you begin to strike it with a hammer, each blow will change the hardness and stiffness of the metal in the vicinity of the blow. If the objective is to achieve a gentle spreading out of the metal, such as might be required for a C-natural key, we will start with a series of blows with the flatter end of the hammer, with proportionally more of them toward the end of the piece.
One of the things I watch for (or, more accurately, listen for) as a guide for how much upsetting the workpiece can take is the sound of the blow. A workpiece’s vibrational characteristics change very quickly as it becomes harder and stiffer. The first blows on a freshly annealed piece will sound dead, as the soft metal can absorb a great deal more of the energy of each blow than it can as it becomes harder. It may take a good few sacrificial pieces to learn how to relate the sound of the blows to the amount of pounding the workpiece can take before it fractures. To be on the safe side, I generally count blows when I begin forging a batch of keys. It is well to remember that there is no good way to repair or recover a piece that has been worked beyond the point at which it fractures, so it is better to err on the conservative side. For the first round of bashing, I will give each piece no more than 15 blows, even if I feel it still sounds soft and could stand more. This small amount of forging does not change the shape of the workpiece a great deal, but every stroke will change it slightly, and the job is to plan the gradual changes of shape of the workpiece so that it morphs gradually into the desired shape. As a general rule, consecutive blows of the hammer should be close together, usually overlapping slightly. I like to position myself with respect to the anvil so that I can see light reflected from the surface being struck with the hammer. The freshly annealed surface will be somewhat matte, but where it has been struck with a polished hammer it becomes more reflective. As with any exercise in shaping something, it is always best to have the desired shape fixed firmly in the mind, so you know where you want to end up.
After the first round of bashing, in my case something like 15 blows per piece, I will anneal all the pieces together. I make a pile of the key forgings in the annealing tray, arranging them so that all of the touch ends, where the heat will be concentrated, are aligned in the same direction. I apply heat to the pile with a torch, always keeping the flame moving over the pile of keys, never dwelling on one piece, but concentrating the heat primarily on the end of each workpiece where it is to be worked. The room will be darkened, and using a pair of long tweezers, I will pull each piece out when it begins to glow, leaving it to cool on another section of the pumice stones. This same procedure will be followed as many times as the forging process requires. In the case of brass chanter keys, this can be as few as four times or as many as fifteen, depending upon the complexity of the key shape. Keys which have bends or curves will typically require more stages.
For the second stage, I will give each piece more blows, typically twenty to twenty-five. I believe it is important to take the initial stages slowly, and not force too much movement of the metal before it is annealed again. I count blows only for the first two or three stages, and once the key begins to take on the recognizable shape of a key, it becomes necessary to pay very careful attention to how each individual piece responds to the hammer.
At this stage, only the flat (slightly domed) end of the hammer head has been used, as this tends to spread the metal out. Starting with the third or fourth annealing, I will begin to use the spherical end of the hammer, as this imparts a slight doming effect to the key touch, giving it something of the shape of a spoon.
A certain amount of work with files and perhaps a disc or belt sanding machine is inevitable with forgings. The rough workpiece should be forged out to a size just a little larger than that of the finished key to allow for the metal that will be lost in the processes of final shaping and finishing. While most of the hammer blows will fall on only one side of the workpiece (i.e. the underside), all sides and aspects of the piece near the blows will change during the process. Depending on how much the metal has been upset and how the blows are directed, the sides of the workpiece may bulge or become concave, and to bring the finished key to a satisfactory regular shape, often far more metal is removed by filing or sanding than one expects. The tuition one must pay to learn what to expect and how to plan for it generally takes the form of a number of failed attempts.
In the next installment, we will take a look at some of the machining and soldering processes used in working up key forgings and begin the examination of the process of key fabrication from sheet and plate stock.
For a complete PDF of the original Summer2012 Pipers Review this article is from, click here.
