It is necessary that a conductor joint shall be mechanically strong and have a relatively low resistance which must remain substantially constant throughout the life of the joint.
Efficient joints in copper busbar conductors can be made very simply by bolting, clamping, riveting, soldering or welding, the first two being used extensively, though copper welding is now more generally available through improvements in welding technology.
Welded joints in copper busbars have the advantage that the current carrying capacity is unimpaired, as the joint is effectively a continuous copper conductor.
Bolted joints are compact, reliable and versatile but have the disadvantage that they necessitate the drilling or punching of holes through the conductors with the bolt holes causing some distortion of the lines of current flow. This joint type also has a somewhat more uneven contact pressure than one using clamp plates.
Clamped joints are easy to make with the full cross-section being unimpaired. The extra mass at the joint and hence cooling area helps to give a cooler running joint and with a well-designed clamp, gives a very even contact pressure. The further added advantage is that of easy erection during installation. A disadvantage is the much higher costs of the clamps and associated fixings.
Riveted joints are efficient if well made, but have the disadvantage that they cannot easily be undone or tightened in service and that they are not so convenient to make from an installation point of view.
Soldered or brazed joints are rarely used for busbars unless they are reinforced with bolts or clamps since heating under short-circuit conditions can make them both mechanically and electrically unsound.
The resistance of a joint is affected mainly by two factors:
a) Streamline effect or spreading resistance Rs, the diversion of the current flow through a joint.
b) The contact resistance or interface resistance of the joint Rj.
The total joint resistance Rj = Rs + Ri.
The above is specifically for a d.c. current. Where a.c. currents are flowing, the changes in resistance due to skin and proximity effects in the joint zone must also be taken into account.
Before considering the effect of the above factors on the efficiency of a joint, it is important to realise the nature of the two contact surfaces. No matter how well a contact surface is polished, the surface is really made up of a large number of peaks and troughs which are readily visible under a microscope. When two surfaces are brought together contact is only made at the peaks, which are subjected to much higher contact pressures than the average joint contact pressure, and hence deform during the jointing process. The actual contact area in the completed joint is much smaller than the total surface area of the joint. It has been shown that in a typical busbar joint surface the effective contact area is confined to the region in which the pressure is applied, i.e., near the bolts in the case of a lapped joint.
Streamline effect
The distortion of the lines of current flow at an overlapping joint between two conductors affects the resistance of the joint. This effect must also occur when the current flows from peak to peak from surface to surface though the overall effect is that through the joint.
In the case of an overlapping joint between two flat copper bars, the streamline effect is dependent only on the ratio of the length of the overlap to the thickness of the bars and not on the width, provided that this dimension is the same for both bars. It has been shown both mathematically and experimentally that even in a perfectly made overlapping joint between two relatively thin flat conductors having a uniform contact resistance, the distribution of current over the contact area is not uniform. Practically all of the current flowing across the contact surfaces is concentrated towards the extremities of the joint and the current density at the ends of the overlapping conductors may be many times that at the centre of the joint.
It is evident from the above that the efficiency of an overlapping joint does not increase as the length of the overlap increases and that from a purely electrical point of view no advantage is to be gained by employing an unduly long overlap.
The relation between the resistance due to streamline effect of an overlapping joint between two flat copper conductors and the ratio of the length of the overlap to the thickness is shown in Figure 15. It has also been found that the distortion effect in a T-joint is about the same as a straight joint.
The resistance ratio e in Figure 15 is the ratio of the resistance of a joint due to streamline effect RS, to the resistance of an equal length of single conductor Rb, i.e.
where a = breadth of bar, mm
b = thickness of bar, mm
l = length of overlap, mm
r = resistivity of the conductor, mW mm
From the graph it can be seen then that the effect falls very rapidly for ratios up to two and then very much more slowly for values up to seven. This means that in most cases the streamline effect has very little effect as the overlap is of necessity much greater than seven.
Figure 15 Streamline effect in overlapping joints
Contact resistance
The contact interface between the two faces of a busbar joint consists of a large number of separate point contacts, the area of which increases as more pressure is applied and the peaks are crushed.
There are two main factors which therefore affect the actual interface resistance of the surfaces.
a) The condition of the surfaces.
b) The total applied pressure.
The type of coating applied to the contact surfaces to prevent or delay the onset of oxidation when operating at elevated temperatures or in a hostile environment is also important, particularly in the long term.
Condition of contact surfaces
The condition of the contact surfaces of a joint has an important bearing on its efficiency. The surfaces of the copper should be flat and clean but need not be polished. Machining is not usually required. Perfectly flat joint faces are not necessary since very good results can in most cases be obtained merely by ensuring that the joint is tight and clean. This is particularly the case where extruded copper bars are used. Where cast copper bars are used, however, machining may be necessary if the joints are to obtain a sufficiently flat contact surface.
Oxides, sulphides and other surface contaminants have, of course, a higher resistance than the base metal. Copper, like all other common metals, readily develops a very thin surface oxide film even at ordinary temperatures when freely exposed to air, although aluminium oxidises much more rapidly, and its oxide has a much higher resistivity.
The negative temperature coefficient of resistance of copper oxide means that the joint conductivity tends to increase with temperature. This does not, of course, mean that a joint can be made without cleaning just prior to jointing to ensure that the oxide layer is thin enough to be easily broken as the contact surface peaks deform when the contact pressure is applied.
Preparation of surfaces
Contact surfaces should be flattened by machining if necessary and thoroughly cleaned. A ground or sand-roughened surface is preferable to a smooth one.
It is important to prevent the re-oxidation of the joint in service and it is therefore recommended that the contact faces should be covered with a thin layer of petroleum jelly immediately after cleaning the contact surfaces. The joint surfaces should then be bolted together, the excess petroleum jelly being pressed out as the contact pressure is applied. The remaining jelly will help to protect the joint from deterioration.
It should be noted that in cases where joints have to perform reliably in higher than normal ambient temperature conditions, it may be advisable to use a high melting point jelly to prevent it from flowing out of the joint, leaving it liable to attack by oxidation and the environment.
The following sections describe the use of coatings on conductor contact surfaces. It should be noted that recent tests carried out to investigate the performance of bolted joints under cyclic heating with wide temperature variations indicate that joints without coatings give the most reliable long-term performance (Jackson 1982). The reason for this is that most coatings are of soft materials which when subjected to continuous pressures and raised temperatures tend to flow. This has the effect of reducing the number of high pressure contact points formed when the joint is newly bolted together.
Tinning. The tinning of the contact surfaces of a bolted or clamped joint with pure tin or a lead-tin alloy is normally unnecessary, although advantages can be gained in certain circumstances.
If the joint faces are very rough, tinning may result in some improvement in efficiency. In most cases, however, its chief virtue lies in the fact that it tends to prevent oxidation and hence subsequent joint deterioration. It may therefore be recommended in cases where the joints operate at unusually high temperatures or current densities or when subjected to corrosive atmospheres.
For the best results the surfaces should be tinned or re-tinned immediately prior to the final joint clamping. It should be noted that both the electrical conductivity and the oxidation protective action decrease as the lead content of the solder increases. Lead also has the effect of reducing the surface hardness of the coating and a high lead content in the tinning material should be avoided as this can cause the plating to creep once the joint is bolted together resulting in premature failure due to overheating.
Silver or nickel plating. This type of plating is being used increasingly, particularly where equipment is manufactured to American standards which require plated joints for high temperature operation. Nickel-plating provides a harder surface than silver and may therefore be preferable. These platings are expensive to apply and must be protected prior to the final jointing process as they are always very thin coatings and can therefore be easily damaged. There is also some doubt as to the stability of these joints under prolonged high temperature cycling. Very high contact resistances can be developed some time after jointing. It is therefore suggested that natural metal joints are in most cases preferable.
Effect of pressure on contact resistance
It has been shown above that the contact resistance is dependent more on the total applied pressure than on the area of contact. If the total applied pressure remains constant and the contact area is varied, as is the case in a switch blade moving between spring loaded contacts, the total contact resistance remains practically constant. This can be expressed by an equation of the form:
where Ri = resistance of the contact
p = total contact pressure
n = exponent between 0.4 and 1
C = a constant
The greater the applied total pressure the lower will be the joint resistance and therefore for high efficiency joints high pressure is usually necessary. This has the advantage that the high pressure helps to prevent deterioration of the joint. Figure 16 shows the effect of pressure on joint resistance.
Figure 16 The effect of pressure on the contact resistance of a joint between two copper conductors
Joint resistance falls rapidly with increasing pressure, but above a pressure of about 15 N/mm2 there is little further improvement. Certain precautions must be observed to ensure that the contact pressure is not unduly high, since it is important that the proof stress of the conductor material or its bolts and clamps is not exceeded.
As a bar heats up under load the contact pressure in a joint made with steel bolts tends to increase because of the difference in expansion coefficients between copper and the steel. It is therefore essential that the initial contact pressure is kept to a such a level that the contact pressure is not excessive when at operating temperature. If the elastic limit of the bar is exceeded the joint will have a reduced contact pressure when it returns to its cold state due to the joint materials having deformed or stretched.
To avoid this, it is helpful to use disc spring washers whose spring rating is chosen to maintain a substantially constant contact pressure under cold and hot working conditions. This type of joint deterioration is very much more likely to happen with soft materials, such as E1E aluminium, where the material elastic limit is low compared with that of high conductivity copper.
Joint efficiency
The efficiency of a joint may be measured in terms of the ratio of the resistance of the portion of the conductor comprising the joint and that of an equal length of straight conductor.
The resistance of a joint, as already mentioned, is made up of two parts, one due to the distortion of lines of current flow and the other to contact resistance. The resistance due to the streamline effect at an overlap joint is given by:
where for a given joint a, b and l are the width, thickness and overlap length, these all being constant, and contact resistance of the joint is:
where Y = contact resistance per unit area.
The total joint resistance is:
and the efficiency of the joint is:
The resistance of an equal length of straight conductor is given by:
The resistance ratio e is obtained from Figure 15.
In most cases it is inadvisable to use contact pressures of less than 7 N/mm2, 10 N/mm2 being preferred. The contact pressure chosen is influenced by the size and number of bolts or clamps, the latter giving a more even contact pressure. For the sake of symmetry the length of overlap is often made equal to the width of the bar, though with thick and narrow bars the overlap can be increased to improve the overall joint performance.
Owing to the larger surface area from which heat may be dissipated, efficient joints between single copper conductors usually have a lower temperature rise than the conductors themselves. It is important, in general, to ensure that all joints have a reasonable margin of safety. This is particularly so where multi-conductors join at one joint and/or the conductors are normally running close to the specified maximum temperature rises.
In deciding the number, size and distribution of bolts required to produce the necessary contact pressure to give high joint efficiency, both electrical and mechanical considerations have to be taken into account. The methods used to determine these requirements have been given in previous sections.
A joint normally decreases in resistance with an increase in the size and number of bolts used. Bolt sizes usually vary from M6 to M20 with between four and six being used in each joint with a preference for four bolts in narrow conductors and six in large conductors. The torque chosen for each bolt size is dependent on the bolt material and the maximum operating temperature expected. Because of the strength of copper, deformation of the conductor under the pressure of the joint is not normally a consideration.
Table 9 shows typical bolting arrangements for various busbar sizes. The recommended torque settings are applicable to high-tensile steel (8.8) or aluminium bronze (CW307G, formerly Cy104) fasteners with unlubricated threads of normal surface finish. In the case of stainless steel bolts, these torque settings may be used, but the threads must be lubricated prior to use.
In addition to the proof or yield stress of the bolt material and the thread characteristics, the correct tightening torque depends on the differential expansion between the bolt and conductor materials. Galvanised steel bolts are normally used but brass or bronze bolts have been used because their coefficients of expansion closely match the copper conductor and hence the contact pressure does not vary widely with operating temperature. Copper alloy bolts also have the advantage that the possibility of dissimilar metal corrosion is avoided. Because these alloys do not have an easily discernible yield stress, however, care has to be taken not to exceed the correct tightening torque.
Because of their non-magnetic properties, copper alloys may also be preferred to mild or high-tensile steel where high magnetic fields are expected. Alternatively, a non-magnetic stainless steel may be used. In most cases however, high-tensile steel is used for its very high yield stress.
Table 9 Typical busbar bolting arrangements (single face overlap)
| Bar width mm | Joint overlap mm | Joint area mm2 | Number of bolts * | Metric bolt size (coarse thread) | Bolt torque Nm | Hole size mm | Washer diameter mm | Washer thickness mm |
| 16 | 32 | 512 | 2 | M6 | 7.2 | 7 | 14 | 1.8 |
| 20 | 40 | 800 | 2 | M6 | 7.2 | 7 | 14 | 1.8 |
| 25 | 60 | 1500 | 2 | M8 | 17 | 10 | 21 | 2.0 |
| 30 | 60 | 1800 | 2 | M8 | 17 | 10 | 21 | 2.0 |
| 40 | 70 | 2800 | 2 | M10 | 28 | 11.5 | 24 | 2.2 |
| 50 | 70 | 3500 | 2 | M12 | 45 | 14 | 28 | 2.7 |
| 60 | 60 | 3600 | 4 | M10 | 28 | 11.5 | 24 | 2.2 |
| 80 | 80 | 6400 | 4 | M12 | 45 | 14 | 28 | 2.7 |
| 100 | 100 | 10000 | 5 | M12 | 45 | 15 | 28 | 2.7 |
| 120 | 120 | 14400 | 5 | M12 | 45 | 15 | 28 | 2.7 |
| 160 | 160 | 25600 | 6 | M16 | 91 | 20 | 28 | 2.7 |
| 200 | 200 | 40000 | 8 | M16 | 91 | 20 | 28 | 2.7 |
* high-tensile steel or aluminium bronze (CW307G, formerly C104)
The choice of clamp material and method of manufacture depends on the a.c. or d.c. current requirements, and on the number of clamps of a given size required. The manufacturing methods used include machining from plate, sand or die casting, or stamping from plate. In the case of low current a.c. (less than 3000 A) and d.c. systems the clamps should be made from a high-strength material compatible with the required contact pressure. They can therefore be made from steel in cast, forged or stamped form. Where a.c. currents in excess of 3000 A are concerned, the choice of material is between the low or non-magnetic steels or a brass or bronze. Steel clamps are generally unsuitable because of the hysteresis losses induced in them.
The inert gas shielded arc processes, tungsten inert gas (TIG) and metal inert gas (MIG) are the preferred welding methods for high conductivity coppers and are capable of producing excellent busbar joints. The welding data given in Table 10 are provided as a guide to good practice, but the actual welding conditions that will give the best results for a particular joint must be determined from experience. Certain physical and metallurgical properties of copper must, however, be taken account of when welding. The high thermal diffusivity of copper - four or five times that of mild steel - opposes the formation of an adequate weld pool necessary for good fusion and deoxidation which can give rise to lack of fusion defects and porosity. The rapid heat sink effect, which is particularly pronounced in thicker sections, must therefore be overcome. Preheating of the copper before welding is necessary for thickness above about 3 mm as indicated in Table 10.
As explained in Section 2, the tough pitch grades of copper, CW004A and CW005A (formerly C101 and C102), contain particles of cuprous oxide which are normally in a form which has a minimal effect on electrical and mechanical properties. Prolonged heating of the copper however, allows the oxide particles to diffuse to grain boundaries where they can seriously affect the properties. This diffusion effect is both time and temperature dependent and is minimised by performing the welding operation as quickly as possible and by restricting the overall heating of the component as far as possible consistent with adequate fusion and a satisfactory weld profile. This consideration obviously does not apply to oxygen-free coppers which do not contain the oxide particles.
Table 10 Welding data for HC copper
a) Recommended usage of BS 2901 filler alloys for TIG and MIG welding of high conductivity copper.
Designation |
Grade |
TIG | MIG |
||
| Argon or Helium | Nitrogen | Argon or Helium | Nitrogen | ||
| CW004A | Electrolytic tough pitch high conductivity | C7, C21 | Not recommended | C7, C8, C21 | Not recommended |
| CW005A | Fire-refined tough pitch high conductivity | C7, C21 | Not recommended | C7, C8, C21 | Not recommended |
| CW008A | Oxygen-free high conductivity | C7, C21 | Not recommended | C7, C21 | Not recommended |
b) Typical operating data for TIG butt welds in high conductivity copper.
(Direct current; electrode negative; argon and helium shielding)
| Shielding gas | ||||||||
| Argon | Helium | |||||||
| Thickness (mm) | Preheat temperature* (°C) | Electrode diameter (mm) | Filler rod diameter (mm) | Gas nozzle diameter (mm) | Weld current (A) | Gas flow (l/min) | Weld current (A) | Gas flow (l/min) |
| 1.5 | None | 1.6-2.4 | 1.6 | 9.5 | 80-130 | 4-6 | 70-90 | 6-10 |
| 3 | None | 2.4-3.2 | 1.6 | 9.5-12 | 120-240 | 4-6 | 180-220 | 6-10 |
| 6 | up to 400 | 3.2-4.8 | 3.2 | 12-18 | 220-350 | 6-8 | 200-240 | 10-15 |
| 12 | 400-600 | 4.8 | 3.2-4.8 | 12-18 | 330-420 | 8-10 | 260-280 | 10-15 |
| >12 | 500-700 | 4.8 | 3.2-4.8 | 12-18 | >400 | 8-10 | 280-320 | 12-20 |
* May be reduced significantly in helium shielding
c) Typical operating data for MIG butt welds in high conductivity copper.
(1.6 mm diameter filler wire; argon shielding)
| Thickness (mm) | Preheat temperature (°C) | Welding current (A) | Arc voltage (V) | Wire feed rate (m/min) | Gas flow rate (l/min) |
| 6 | None | 240-320 | 25-28 | 6.5-8.0 | 10-15 |
| 12 | up to 500 | 320-380 | 26-30 | 5.5-6.5 | 10-15 |
| 18 | up to 500 | 340-400 | 28-32 | 5.5-6.5 | 12-17 |
| 24 | up to 700 | 340-420 | 28-32 | 5.5-6.5 | 14-20 |
| >24 | up to 700 | 340-460 | 28-32 | 5.5-6.5 | 14-20 |
Thermal expansion should be allowed for during welding as this leads to the closing of root gaps as the temperature of the metal rises. The root gaps indicated in Table 11 should therefore be allowed.
Oxy-acetylene and oxy-propane welding methods can be used with oxygen-free copper but they are not recommended for welding tough pitch coppers as the reducing atmosphere produced in the flame can react with the cuprous oxide particles to produce steam inside the metal. This gives rise to porosity and is known as 'hydrogen embrittlement'.
Further details of the factors involved in the welding of copper can be found in the CDA publication No 98.
Table 11 Recommended edge preparations for TIG and MIG butt-welds.
