Copper-Nickel Welding and Fabrication

The following information is adapted from a CDA Publication of the same name published by CDA Inc. as A7020-99/13. Nickel Institute as 12014 Second Edition and CDA UK as Publication 139 Second Edition 

General Handling

The precautions required for handling copper-nickels will be familiar to any fabricator who routinely handles materials like stainless steels and aluminium alloys, but may be new to those used to dealing with only carbon steels.

Cleanliness is most important as contamination can cause cracking and porosity during heat treatment or welding and may affect the corrosion resistance of the alloy. Ideally fabrication should take place in an area devoted solely to Cu-Ni alloys. Where this is impracticable, the standard of care of the material should be well above that necessary for carbon steels.

• Sheets should remain in their packing until needed and should be separated by protective material or other means to avoid abrasion.
• Plates and sheets are best stored vertically in racks which have the steel frames covered.
• Walking over sheets should be avoided.
• Plastic film may be interposed between the sheet and rolls when roll forming.
• Grease and paint should be kept away from the surface particularly near edges of weld preparations, while all trace of marking crayons must be removed before making a joint.
• Stainless steel brushes should be used and tools such as grinding discs should not be used on Cu-Ni alloys if they have been used on other materials.
• Openings of pipes and fittings must be protected on completion of fabrication to prevent ingress of dirt etc. before installation.

Cutting and Machining

Most normal cutting processes are acceptable for Cu-Ni such as shearing, abrasive disc cutting and plasma arc. High-speed abrasive wheels work well for bevelling edges and trimming material. Laser and abrasive water jet cutting are also possible.

Oxy-acetylene cutting is not appropriate for these materials. Band saws or shears may be used for cutting, but allowance must be made for the fact the alloys are relatively soft and ductile.

Although Cu-Nis are not as readily machined as free cutting materials such as brass, they are not difficult to machine and can be ranked along with aluminium bronze and phosphor bronze alloys. They are much easier to machine than say stainless steels and other alloys which work-harden rapidly.

More details and recommended speeds and oils are detailed in:

Machining Brass, Copper and its Alloys, CDA Publication TN 44.

Recommended Machining parameters for Copper and Copper Alloys; DKI Monograph [i018e].

Forming

Cu-Nis can be hot and cold formed although cold working is preferred. If cold forming is used a full inter-stage anneal may become necessary when the amount of cold work exceeds about 40-50%. A 20% cold reduction approximately halves the as -annealed elongation and doubles the proof strength.

Tubes can be bent by a range of methods, including rotary draw bending, 3-roll bending, compression bending and ram bending (press bending). When bending copper-nickel, a mandrel and wiper die are also applied for support (mandrel bending). Care must be taken to get smooth bends and avoid wrinkling, because liquid turbulence in service can lead to impingement attack. Bends with a tube bend radius of twice the tube diameter can be produced. Smaller radii require prefabricated bends.

More details on bending are given in the Tube and Pipe Bending of Copper Nickel Section.

Hot working Cu-Nis can lead to hot cracking and therefore should be avoided or only attempted with advice from a supplier. The temperature ranges are:

Heat Treatments

The work-piece should be clean and free from any contamination before and during heating.
Cu-Nis can embrittle if heated in the presence of contaminants such as sulfur, lead, phosphorus and other low melting point metals. Sources of contamination include paints, marking crayons, lubricating grease and fluids, and fuels. Fuels used must be low in sulfur; normally, fuel oils containing less than 0.5% by weight sulfur are satisfactory.

Oxidising atmospheres will cause surface scaling. Furnace atmospheres should be neutral to slightly reducing and must not fluctuate between oxidising and reducing conditions. Flame impingement must be avoided.

For a full anneal, soaking times of 3-5 minutes per mm thickness can be used. Recommended temperatures are:

Stress relieving is seldom used but if required the recommended temperatures are:

Descaling

The surface oxide films on both alloys can be very tenacious. Oxides and discolouration adjacent to welds can be removed with very fine abrasive belts or discs. If pickling is required, a hot 5-10% sulfuric acid solution containing 0.35g/l potassium dichromate is satisfactory. Before pickling, oxides can be broken up by a grit blast. The pickled components should be rinsed thoroughly in hot, fresh water and finally dried in hot air.

Welding

Copper-nickels can be readily welded by all conventional processes and, since they have a simple metallurgical structure, do not require preheat or post-weld heat treatment. However, it is essential that requirements for preparation, particularly cleanliness, are carefully followed and that welders undergo a period of familiarisation with the particular characteristics of these alloys if they are not to encounter problems. Automatic welding, including orbital welding of pipe, may also be appropriate.

In some applications, insurance and inspection bodies may require qualification of both welders and welding procedures to appropriate standards. A welding procedure specification (WPS) should be prepared in all cases.

Since the predominant application of copper-nickels is in the form of relatively thin-walled pipe, the gas-shielded tungsten-arc welding process (known as TIG or GTAW) is frequently used, both for joining pipe sections and for attaching fittings and flanges.

The most widely available welding process is the manual metal arc process(known as MMA or SMAW) using flux coated stick electrodes. This is quite suitable for welding Cu-Ni alloys and has the advantage of using relatively inexpensive equipment.

For thicker materials, above 6 mm, the TIG (GTAW) process can be used for the root run before completion of the weld with by the MMA (SMAW) process. The gas-shielded metal arc process (known as MIG or GMAW), using a continuous wire feed, is faster and can be closely controlled with modern sophisticated equipment.

Mechanical Properties of Welds

A copper-30% nickel filler material is recommended for the welding of the 90-10 and 70-30 Cu-Ni alloys. Because of the higher nickel content, the weld metal is stronger than the 90-10 Cu-Ni base metal. In making a transverse tensile weld test in a 90-10 Cu-Ni weld qualification plate test, all the elongation can be concentrated in the heat affected zone if the specimen moves. This can be prevented by using a longitudinal bend test specimen instead. Typical properties of the all weld metal are shown in Table 10:

Preparation for Welding

If stored correctly, material to be welded should be in a generally clean condition. Dirt of any kind must be removed along with residual oil and grease. Particular attention should be given to sources of the elements that can cause cracking or microfissuring in the weld, which can originate from crayon or paint identification markings, temperature indication markers, and other contaminants. (Fittings of other alloys, such as gunmetal - copper-tin-zinc alloy - are also a source of detrimental elements and should not be welded to copper-nickel alloys.)

The joint area should be thoroughly cleaned before welding starts. Particular attention should be paid to the weld preparation and an adjacent area at least 10mm wide, preferably wider, either side of the preparation, which can be degreased with an uncontaminated organic solvent applied with a fine abrasive pad or a clean cloth. The area should be dried with clean cloths. Their appearance after use is an indicator of cleanliness: they should be free from any residue.

Weld Preparations

It is possible to weld copper-nickel up to 3mm thick with a square butt preparation. However, autogenous welding should not be attempted, since this will result in porous welds due to the absence of effective deoxidisers in the alloys. Above this thickness, a bevelled preparation must be used; the included angle of the V should be larger than for carbon steel - typically, 70° or more - because the molten weld metal is not as fluid as with carbon steels, and manipulation of the electrode or torch is necessary to ensure fusion with the side walls.

Although it is possible to weld in all customary welding positions, it is desirable to weld down-hand, which allows higher deposition rates and may demand less skill. It will often be impracticable to turn large or complex structures into this most favourable position for welding, but it is worth the effort of manipulating subassemblies for down-hand welding, where possible.

There is no need to pre-heat the base metal before tacking or welding unless this is necessary to ensure that the base metal is dry. To avoid microfissuring, the interpass temperature is maintained below 150°C.

Tack Welding

Because of their high coefficient of thermal expansion relative to carbon steel, Cu-Ni's have a greater potential for distortion when welded. Welding fixtures can help but their use is limited to sub-assemblies. Tack welds should therefore be made to maintain a uniform gap and alignment between the parts being welded. They must be positioned at about half the spacing usual for carbon steel and are preferably quite short. The TIG (GTAW)process is often used for tacking, although, where the equipment has the facility available, the MIG (GMAW)spot-welding process is a convenient and well controlled technique for the purpose. Tacks should be wire brushed or ground to clean metal where they are to be incorporated into the weld metal of the joint.

Welding Consumables

While consumables are available that deposit weld metal similar in composition to the 90-10 copper-nickel alloy, welds made with them may not have adequate corrosion resistance for all applications. Consumables for the 70-30 alloy, on the other hand, offer superior deposition characteristics and the corrosion resistance of 70-30 weld metal is at least comparable to each of the base metal alloys. These consumables are therefore recommended for both types of alloy.

For welding copper-nickel to steel, nickel-copper consumables containing about 65% Ni are used as the weld metal can absorb more iron dilution from the steel without cracking than copper-nickel weld metals.

Many weld consumable manufacturers offer Cu-Ni and nickel-copper electrodes and filler wires to recognised specifications, Table 11. These contain additions of titanium and manganese to react with nitrogen and oxygen from the atmosphere which would otherwise create porosity. If weld metal porosity persists despite the use of the correct filler material, the most likely causes are inadequate shielding of the weld pool and improper weld joint cleaning. Other possible causes include an excessively long arc, moisture on the weld preparation or the use of coated electrodes which are not fully dry.

For the Manual Metal Arc (MMA or SMAW) Process:

• Flux-coated electrodes are designed to operate with direct current, electrode positive.
• No special electrode baking or drying treatment is required unless they have been exposed to the atmosphere for some time. In this case, they should be dried in an oven, e.g. for 1-2 hours at 250C.
• An electrode size slightly smaller than that of a carbon steel electrode under comparable conditions is preferred taking into account the need for manipulation.
• Any weaving should not be more than three times the electrode diameter.
• A long arc should be avoided, since this results in weld porosity through reaction with the surrounding atmosphere.
• Start positions can be unsound and reversing the electrode direction to remelt initially deposited weld metal or the crater at the end of a run can help to avoid problems.
• Slag must be removed between runs by chipping and brushing to leave a clean surface for the next run.

For the Gas-shielded Tungsten Arc (TIG or GTAW) process:

Compared to MMA (SMAW), separate control of heat input via the arc and filler material addition gives TIG (GTAW) a degree of flexibility which is an advantage when welding shaped joints. In general, the process is suitable for joining materials up to 6 mm thick and is the method commonly used for welding thin walled pipes. It is also preferred for tacking and for inserting root runs in thicker joints that are completed by the MMA process with flux-coated electrodes. Automatic equipment is available for orbital welding of pipe and other applications.

To offset a greater risk of weld metal porosity than with the other processes, the weld pool must be protected as far as possible from contact with the atmosphere by maintenance of a short arc and adoption of a stringer bead technique; weaving of the torch is undesirable. It is essential that the weld pool is fully deoxidised by addition of filler metal throughout a run. Autogenous welds are very likely to be porous. If the filler metal is accidentally withdrawn at any point, that part of the weld should be ground out and repaired.

Argon is recommended as the shielding gas. Gas flow should be maintained at the end of a run until the weld pool has solidified; crater filling devices are beneficial. The interior of pipes should be purged with argon prior to and during welding. Where joints are made with backing bars, these should be made of copper or copper-nickel alloy.

Direct current should be used.

For the Gas-shielded Metal Arc (MIG or GMAW) process:

Due to the higher capital cost of equipment and the necessity of buying spools of filler wire, MIG (GMAW) is more appropriate for extensive welding operations.

MIG (GMAW) can be operated over a range of currents to provide various transfer modes:

Dip (or short circuiting) transfer

- low heat-input and used for thinner sections

Spray transfer

- relatively high heat-input and only suitable for thicker materials, say above 6 mm thickness and down hand welding

Pulsed-arc transfer

- a technique in which metal transfer is closely controlled providing a combination of low overall heat input and adequate fusion to the base metal. It is suitable for a range of thicknesses.

Because of the range of transfer conditions which are possible with the gas-shielded metal arc process, welding parameters can vary widely. In all cases, these should be set for the equipment and position and thickness of the material by careful welding procedure trials directed towards stable transfer conditions and welds of good appearance.

• Argon or a mixture of argon and helium is preferred as shielding gas.
• The spooled filler wire must be kept dry and not exposed to contamination.
• Attention should be paid to the effectiveness of the wire feeding system when welds have to be made some distance from the welding equipment since filler wire is relatively soft.
• Low friction liners are essential for the feed hose.

Post-weld treatment

No heat treatment is necessary after welding. All traces of slag should be removed from joints made by the manual metal-arc process and the weld area may be cleaned e.g. with a rotating flap wheel or stainless steel brush, to leave a bright finish.

Inspection

Welds should be inspected visually for defects such as cracks, undercut, lack of fusion and penetration, and weld contour. Liquid dye penetrant inspection is a simple method for ensuring that there is no cracking at the surface. For critical applications, more advanced inspection techniques are adopted, such as radiography, but these are not required for general fabrications.

Clad Plate-Preparation and Welding

An economical way of constructing thicker plate section can be to use steel plate which has been roll-clad with 90-10 or 70-30 Cu-Ni alloy. Examples are tube sheets and water boxes. Also 8mm (2mm copper- nickel and 6mm steel) thick plate has been successfully used to build four fire boats in Italy. This type of material can offer significant advantages in some situations but is not so readily available as the solid Cu-Ni alloy itself.

Clad plate should be handled with special care appropriate to the Cu-Ni alloy and not treated as normal structural steel.

Unlike solid Cu-Ni plate, it is possible to use oxyacetylene equipment for cutting clad plate if the ratio of steel to clad thickness is 4 to 1 or greater (20% clad or less). The clad side of the plate should face downwards so that cutting is initiated from the steel side to allow the slag stream from the backing steel to act as a cutting agent for the cladding. This precaution is not necessary for plasma-arc cutting but some trials may be required to find the most suitable settings for either cutting procedure. It is essential that the cut face is ground or machined to clean metal in forming the weld preparation.

When designing weld procedures for clad plate, it is necessary to treat the cladding and backing material as separate components and to avoid the respective weld metals being mixed. Otherwise, cracking is likely to occur from copper in carbon steel weld metal or iron in Cu-Ni weld metal. The region adjacent to the interface between the backing material and the cladding is welded with the 65% nickel-copper filler material which can cope with iron pick-up from the carbon steel side. When the clad thickness is about 10mm or less, the 65% nickel-copper filler metal is often used for the complete weld.

When it is possible to weld from either side, the steel side is welded first and the assembly is then inverted. The cladding is prepared for welding, cutting into the steel weld and allowing for at least two runs; the first being of the 65% nickel -copper alloy referred to earlier followed by the 70-30 Cu-Ni.

When access is only possible from the steel side, the joint is prepared to leave the copper- nickel cladding protruding, so that it can be welded the same as for the solid alloy. The weld joint in the steel backing is then made with the 65% nickel-copper followed by the steel filler runs.

If the preparation is made from the cladding side, the joint is partially filled with steel weld metal and then completed with the combination of the 65% nickel-copper filler and then 70-30 Cu-Ni filler.

Linings

Lining a vessel with Cu-Ni sheet can be a convenient and economic alternative to the use of the solid alloy or clad plate. An early example was the construction of a water box in which the lining was fabricated as a separate component from 1.2mm thick 90-10 Cu-Ni sheet, made to fit closely into a carbon steel shell. It was then attached to the shell by a pattern of MIG (GMAW) spot welds, using an automatically timed sequence. It was necessary in this case to ensure that the lining fitted closely in the shell and was in intimate contact with it when the welds were made. Seal welds made around the flanged opening completed the lining process. Automatic spot welding allowed welds to be made with a 70-30 Cu-Ni alloy filler wire with reproducibly low iron dilution.

In recent years, lining techniques have been extensively developed for lining vessels and ducting with corrosion-resistant alloys, particularly in the power generation industry. Usually, spot welds are used to minimise bulging due to differences of thermal expansion between the backing material and the lining or from the pressure variations and the lining is attached as sheets or strips by a carefully designed welding procedure. It is important that the backing material surface is thoroughly cleaned, e.g. by grinding and blast cleaning with abrasives to produce an uncontaminated surface. The final surface should be closely inspected and any areas of localised thinning must be repaired before commencement of lining.

Two welding procedures are commonly adopted for lining:

• In the first, each sheet is fillet welded to the backing material and then a third, covering bead is deposited to complete the joint.
• In the second procedure, each strip is tack welded to the backing material, overlapping the adjacent sheet by a few centimetres. A seal weld is then made directly between the strips.

With both procedures, it is advisable to use the 65% nickel-copper filler material, although the 70-30 Cu-Ni filler can be used for the seal weld in the second procedure.

The number and pattern of spot welds is determined by the area of sheet or strip between the welds. The reproducibility of the technique also makes it ideal for the repetitive sequence of tack welds. Fillet and seal welds are best made by the MIG (GMAW) process since it operates at relatively high speeds and can be closely controlled by modern power sources. Details and regions of complex shape may be welded by the TIG (GTAW) process which, although slow, is flexible and facilitates manipulation of the torch by the welder.

Throughout fabrication of a lining, care must be taken to avoid surface damage to the Cu-Ni sheet and, on completion, any weld spatter and discoloration must be removed. Welds should be examined visually for defects and the absence of porosity or cracks breaking the surface of welds can be confirmed by a penetrant inspection technique.

Brazing

Copper-nickel alloys are readily brazed by all processes although torch brazing is more common. Since the process relies on wetting of the surfaces to be joined by the brazing alloy, absolute cleanliness is essential. Fluxes alone are not capable of removing all contamination, particularly those containing lead or sulfur, and oils, paint etc. which should be removed carefully with solvents and degreasing agents. Oxides and dirt can be eliminated with emery paper or a chemical cleaning process.

If parts have been cold formed, they may contain significant internal stresses, which promote intergranular penetration by molten filler material during brazing, resulting in cracking at the joint. Removal of stresses by full annealing is not necessary; heating to 600-650°C for a few minutes is sufficient for adequate stress relief and this can be done simply with an oxy-fuel torch, taking care that the part is heated uniformly.

While phosphorus-bearing brazing alloys are often recommended for joining copper alloys, they are not suitable for copper-nickels because the nickel reacts with phosphorus to form a brittle nickel phosphide phase. Silver-based brazing alloys (‘silver solders’) should be used. They offer a useful combination of melting range, flow characteristics and mechanical properties. They also perform well in brazed joints where copper-nickels are exposed to seawater. Alloys containing cadmium are no longer recommended because of health hazards in application, but there is a range of silver-copper-zinc alloys which are suitable and safe.

For brazing pipe and fittings, pre-placed brazing alloy rings are preferred over manual feeding, providing better control of quality and minimising the use of flux, residues of which must always be removed after the joint has been made, usually by washing with hot water. The larger the pipe size, the more difficult it is to achieve uniform heating around the diameter to reach the brazing temperatures. Some organisations limit brazing to pipe diameters up to and including about 50mm.

Furnace brazing is possible and advantageous where significant numbers of assemblies are to be joined. Exothermic, endothermic or dissociated ammonia atmospheres are suitable, together with inert gas because of the high vapour pressure of some brazing alloy constituents, vacuum brazing is less suitable.

Painting

Although painting of Cu-Ni is seldom required as the alloys already have inherent corrosion and biofouling resisting properties, there are occasions when painting is desirable e.g. for aesthetic reasons or to reduce the exposed metal area in a bimetallic couple and reduce the risk of galvanic corrosion.

Cu-Ni can be painted. A thorough roughening by grit or sand blasting is all important before paint is applied. Above the water line on boat hulls, appropriate epoxy followed polyurethane coatings can be applied. Leading paint suppliers will normally prefer to recommend appropriate paint specifications based on their proprietary products for specific applications.