Selecting filler metals for copper and its alloys

TWI Bulletin, July/August 1991

Mike Scott graduated from Cambridge with a degree in natural sciences, specialising in metallurgy. In his first post, with the Royal Naval Scientific Service, he worked on electroplating applied to manufacturing radar valves and ancillary devices, and on the manufacture of ceramic ferrite.

From there he moved to the Central Electricity Research Laboratory where he worked on the metallurgy of magnesium alloys used for sheathing uranium fuel elements for nuclear power stations. In 1960, he joined the then British Welding Research Association, where he spent many years as leader of the Non-Ferrous Group in the Materials Department. He is now a consultant within that Department. During his time at Abington, he has worked on various aspects of joining most of the common non-ferrous metals and their alloys and has been much involved in microjoining.

Last year saw the publication of a revised edition of BS 2901 on filler rods and wires for gas shielded arc welding. In their sequel to selecting aluminium filler metals Mike Gittos and Mike Scott look at filler materials for copper and its alloys.

It is rather more difficult to discuss filler metals for copper and its alloys than it was for aluminium in the authors' previous article ^[1] since there is a much greater range of the former materials and a correspondingly greater range in properties. Nor is there any commonality of welding problems. For example, in welding pure copper, by far the greatest problem is the high thermal conductivity, but this does not apply to copper alloys.

Another of the difficulties in discussing copper is that there is no widely accepted alloy designation system. (There will undoubtedly be such a system in the European Standards for parent metal but, although work has now begun on these standards, the designation system is not yet available.) The code letters used hereafter are therefore basically taken from current British Standards. Table 1 lists them and gives a brief description of each group.

Table 1 Alloy groups (wrought and cast)

Some more detailed discussion of class CH is relevant since it contains very different alloys. One of these is relatively old, the others are much newer. The old one is cap copper (originally used for percussion caps) which contains 5%Zn and is weldable. So is an alloy containing 3%Si-1%Mn, although it is seldom encountered. There are several alloys containing beryllium which can usually be welded but seldom are nowadays, because of health fears. There is an alloy containing 1%Cr (which combines high strength and good conductivity) which can be arc welded but only with care and in situations of low restraint. Electron beam welding has been used successfully.

Filler metals

Table 2 lists the filler metals specified in BS 2901 Part 3, together with their nominal compositions. They have been grouped together rather than presented in numerical order as in Part 3. Simplistically, C1A, C7, C8 and C24 are filler metals for copper; C10, C11 and C27 are for bronzes; C12, C12Fe, C28, C29, C13, C20, C26, C22 and C23 are for aluminium bronzes; and C16 and C18 are for cupro-nickels. However, in many cases, the filler metals in a given group are not interchangeable.

Table 2 Cu-alloy filler metals to BS2901: Part 3:1990

About the only feature the filler metals have in common is that they contain a deoxidant, C1A being the only exception. This deoxidant may be present as an addition specifically for that purpose, e.g. the titanium in C16 and C18, or present primarily as an alloying element that also confers those properties that make the parent metal attractive, e.g. aluminium in the various aluminium bronze filler metals. Although essential, the presence of deoxidant in the filler metal for copper itself is a drawback, since a side effect is a reduction in conductivity, the principal reason for using copper in many cases. However, concern about this aspect is usually greater than it need be, since published information is based on the composition of the filler metal whereas what really matters is the composition of the weld metal. Usually, the latter contains less deoxidant than the former because the deoxidant forms an insoluble oxide in doing its job and this oxide has negligible effect on conductivity.

Table 3 suggests filler metals for welding the classes of copper alloy given in Table 1, both to themselves and to each other. Because of the wide range of alloys covered, additional comments are necessary in some cases; these are too long to be presented as notes to the Table and are given below.

Table 3 Choice of filler metal

Welding

For class C, welding is usually carried out with argon and/or helium shielding, helium offering the considerable advantage of a hotter arc and thus requiring less preheat. Nitrogen shielding gives an even hotter arc but at the expense of a rather dirty looking weld and some difficulty in welding other than in the flat position. It may also require the more strongly deoxidised C8 filler metal. For welding phosphorus-deoxidised copper, C1A can be used if the parent metal is thin, preheat low and the weld pool small.

For class CA, the less highly alloyed filler metals C12, C12Fe, C28 and C29 are used for welding single phase parent metals of similar composition. The more highly alloyed C20 and C26 are used for two phase parent metals, again of basically similar composition. C13 can be used for most parent metals and has the not altogether accurate reputation of being easier to use. However, it suffers from poorer corrosion resistance and where corrosion is likely its use is best avoided. One way round this is to use a capping run over the C13, but cases have been known where subsequent dressing has exposed the C13 weld metal, with unhappy consequences! C22 and C23 are used for parent metals of matching composition.

For classes PB and G, the tin content of the parent metal is usually a good guide as to which of the available filler metals should be used. As indicated in Table 1, leaded alloys are not usually weldable but if the lead content is low (less than 3%, say) C25 may be satisfactory for making cosmetic repairs.

Although autogenous welding of thin sections can be successful, generally classes CZ and NS are best not arc welded, because the zinc evaporates causing porosity and, on occasions, a surprising amount of cobwebby oxide over the work, the torch and the jig. A filler metal with a strong deoxidant is the only choice if welding must be carried out: such filler metals work by forming a thin film of oxide over the weld pool which restricts evolution of zinc vapour. C9 is the best choice, followed by the aluminium bronzes, but a phosphor bronze filler metal may be needed for some dissimilar welds.

For Class CN, 90/10 parent metal can be welded with C16 or C18 filler metals, but 70/30 parent metal can only be welded with C18 (C16 weld metal is commonly anodic to 70/30 parent metal in service). It is therefore very common to use C18 irrespective of the parent metal. C16 and C18 have little tolerance to pick-up of aluminium before becoming embrittled whereas aluminium bronze filler metals have a greater tolerance to pick-up of nickel. For a CA to CN weld, an aluminium bronze filler metal should therefore be used.

This article has covered filler metals for inert gas shielded arc welding processes (TIG, MIG, plasma). In a limited number of cases, manual metal arc electrodes of similar composition to the filler metals are commercially available. The most important class is probably the cupro-nickels which are quite widely used, often after a TIG root run. Phosphor bronze and aluminium bronze electrodes are also available, although the running characteristics of the latter are not as good. None of these consumables is covered by a British Standard, although they may well be covered in the next few years by a European Standard.

In so short an article it is not possible to cover every alloy or alloy combination in detail. In any case of doubt or difficulty, TWI is always pleased to advise member companies.

Reference