Electrical connection

3,002,173

Sept. 26, 1961

R. K, ALLEN

ELECTRICAL CONNECTION

<img file="US3002173A_D0001.tif" />

<img file="US3002173A_D0002.tif" />

<img file="US3002173A_D0003.tif" />

/ 2 3 4 5 6 7 8 8 iO // 12 i3 14 76 7! AT

RAT/0 OF MAJOR AXIS TO DIAMETER OF CIRCLE OF SAME AREAIr?\ren torr Robert /&lt;. Aller/, by -—

Atberney.

3,002,173

Patented Sept. 26, 1961

United States Patent Office

3,002,173

ELECTRICAL CONNECTION

Robert Kilgannon Allen, Erie, Pa., assignor to General Electric Company, a corporation of New York

Filed June 25, 1957, Ser. No. 667,829

Claims. (Cl. 339—95)

My invention relates to a bus bar assembly and, more particularly, to a conductive metal joint in a bus bar assembly wherein molecular electrical contact is assured.

Within the portion of the electrical industry requiring high current capacity bus bars, a means of connecting bus bars with a reliable low resistance, high capacity, detachable joint has been an economic problem for many years. Except for the noble metals, all conductors have oxides on their surfaces. Because of its low cost and relatively good conductivity, copper is often used for high current capacity bus bars despite the occurrence of oxides and resistance films which inhibit current flow between the contacting surfaces of copper through a supposedly low resistance joint. This is particularly troublesome in corrosive or high temperature applications. With a contact joint of two inch bus bars having a two inch overlap, it has been the practice of the manufacturing industry to prepare the surfaces to be connected by removing all film before clamping the surfaces together with a high tension bolt. This rather elaborate and expensive procedure results in an apparent contact area of approximately four inches square. However, the actual molecular contact is substantially less than this. Moreover, at points of no contact, a tarnish will form, particularly at elevated temperatures. When these tarnishes do form, they tend to counteract the bolt tension with a non-conductive layer and thus reduce the support at the conducting surfaces and inhibit current flow to cause an increase in losses and result in additional heating of the joint.

In order to understand completely the scope of my invention, it is necessary to understand the theory of electrical contact joints. In a joint contact, substantially all of the electric current passes from one conductor to the next at points of electrical contact between the conducting metals. Generally speaking, electrical contact is found only at points of metal contact which may be envisioned as direct contact between a conductive atom of one conductor with a conductive atom of another conductor. If one of the contacting atoms is non-conductive, no current will flow except for “tunnel effect” where an electron will flow through a very short space between the conductive metals when the space is no greater than a few angstroms.

In view of the fact that the competitive cost of making a conductor prevents manufacture of surfaces that are flat enough and smooth enough to provide molecular contacts or tunnel effect conductance over the entire “apparent contacting” surface, it becomes obvious that with large contact areas, only high points or peaks (of a few microns in height) actually make electrical contact with an adjacent conductor. Since non-conductive films and oxides are often as thick as the high spots, or hard enough to deform the peaks without being ruptured sufficiently to allow either contact or tunnel effect conductance, it has been the practice to clean carefully the contact surfaces.

Even with carefully cleaned surfaces, the joint resistance is not predictable immediately because of the phenomena of constriction resistance occurring at each contact peak. Constriction resistance may be explained most easily as the effect of contact of the metal at the high points causing a concentration of all the contact current at these very small points. The constriction resistance is a maximum in small circular contact and is inversely proportional to the radius of the contact. However, if the contact point is elliptical, the constriction is reduced.

The constriction resistance of a long contact area having a major axis 28 times as great as that of a circle of the same area is approximately 20 percent as great as that of the same area in a circular form. Understanding of the 5 nature of contact resistance will facilitate comprehension of the following discussion.

It invariably happens that from time to time such conductive metal joints must be disassembled and reassembled in the field (to replace one of the bus bars or to 10 repair insulation, etc.) where proper surface preparation is not practical. With the tarnish accumulated on the surfaces during use prior to the repair in the field, a large percentage of the stress of the clamping bolt is absorbed by what were formerly contact peaks engaging a non15 conductive film or tarnish. Thus, resistance of the reassembled joint is increased and overheating damage is more likely to occur. This has proved particularly expensive within the locomotive equipment history, because a failure of electrical joints in service necessitates the re20 moval of the locomotive to a repair area and often causes the delay of an entire train. One attempt to prevent such failures has been to tinplate the mating surfaces of the joints to provide a soft surface which will deform to provide a greater bearing area and will rupture easily 25 any surface oxides. However, this is very expensive in view of the fact that a very small percent of the joints are subject to such failure and, furthermore, it does not solve directly the problem but merely removes the likelihood of the formation of permanent non-conducting oxides and 30 surface films.

It is therefore a primary object of this invention to provide an economical, simple and reliable electrical connection, suitable for connecting heavy duty bus bars.

Briefly, in carrying out my invention in one of its modi35 fications, a joint for a high current capacity bus bar is formed of two overlapping bus bars which are deformed to reduce the apparent contact area between them and thus increase the actual electrical contact. The contact area is particularly designed to provide a sliding engage40 ment and a sufficient pressure between the contact surfaces under pressure to puncture any non-conducting surface films. The deformation is arranged to provide a non-circular conducting area to minimize constriction resistance.

₄₅ The subject matter which I regard as my invention is particularly pointed out and distinctly claimed in the concluding portion of this specification. My invention, however, as to its organization and method of operation together with further objects and advantages thereof will 5θ best be understood by reference to the following description taken in connection with the accompanying drawing, in which:

FIG. 1 is a top plan view of my invention showing a bus bar provided with male and female deformations;

FIG. 2 is a side plan view of the bus bar shown in FIG. 1;

FIG. 3 is an enlarged detail section view of an assembled joint;

FIG. 4 is a sectional view of a modification of my in66 vention; and

FIG. 5 is a curve of the effect of the shape of a small contact area on constriction resistance.

Referring now to the drawing, in which like numerals refer to similar parts, in FIG, 1, I have shown a bus bar _6g 10 provided with apertures 11 and 12 suitable for accommodating clamping bolts. Coaxially surrounding the aperture 11,1 have provided an annular groove or recess 14. A coaxial annular ridge 15 surrounds the aperture 12 and is adapted to fit into the similar recess 14 of an ad76 jacent bus bar. The annular ridge 15 is semi-toroidal in shape and rises substantially above the flat surface of the bus bar 10.

3,002.173 the ridge 15 and recess 14 wherein a punch 25 is forced into the bus bar 10 at a pressure depending on the thickness and composition thereof. For quarter inch copper and a one inch diameter ridge, the punch pressure should 5 be about 30 tons. A mating female mold (not shown) controls the contour and diameter of the annular ridge 15. A second punch 26 is then used to form the concave surfaces 19 . and 20. The critical feature of the molds and punches is that the surfaces of contact should 10 hot provide a wide area but should instead approach a sharp line contact as shown at the upper corner of the surface 19. Thus, the sloping edge of the ridge 15 engages only the peripheral edge of the groove 14 with a wiping action.

Although this is one of the most economical methods of producing my invention on bus bars where mass production techniques may be used, my invention may be modified for use on single applications by machining the contours as shown in FIG. 4. In this modification, 20 a female groove 30 is cut from a bus bar 31α coaxially with a clamping bolt aperture 32, and in a bus bar 31ά mating annular ridge 34 remains coaxial with a bolt aperture 35 after cutting away the surrounding copper surface. In this modification, the annular ridge 34 has a 25 triangular cross-section. As shown in FIG. 4, the annular ridge 34 in a bus bar 31b is inserted in the groove 30 of a bus bar 3ία. I prefer to have the annular groove 30 designed so that the outer acute comers 36 of the groove 3® engage tightly the outer, or the inner, surfaces 30 of the ridge 34. Because of the acute angle Θ,. this modification insures a very good wiping action to destroy tarnish and reduce constriction resistance.

According to my invention, the criteria are to provide a small apparent area of mechanical contact which is 35 the actual electrical contact area, to provide sufficient mechanical pressure to achieve a good wiping action and substantially complete rupture of high resistance oxides or films covering these areas, and to provide a long (perimeter) area to reduce constriction resistance.

4.0 With such a small contact area, where the apparent area is of the order, of one percent of the nominal crosssection area of the bus bar, there appears to be a problem of excessive heating with high current capacity bus bar joints. Obviously, the current within this very small conducting area is much higher density than that normally ⁴³ allowed in conductors. However, I have found that current density as high as 57,000 amperes per square inch does not show appreciable saturation effect.

After considerable reflection on this phenomena, I have reached the conclusion that there is little relation³³ ship between the apparent area of contact and actual electrical contact in conventional bus bar joints, and that, because of current concentration at the contacting peaks, the current density in a bus bar joint point of contact is actually much greater than before calculated. This is ⁵⁵ actually of little consequence, even though it can be assumed that the temperature rise of very small contact points would cause high resistance at that point, because the heat generated is small compared to the amount of thermally conductive metal surrounding it. Thus, al⁶⁰ though the temperature of the area of contact is high, and its resistance is thereby increased, the depth of the high temperature is so slight that it does not materially affect the overall resistance of the joint. My invention takes advantage of this phenomena by the use of a rela⁶⁵ tively small contact area located between relatively large thermally conductive metal masses.

Moreover, since the contact pressure and wiping action of the mating surface 19 or 20 is enough to destroy ₇₀ all films and fracture many of the peaks, I provide an increased metal to metal contact by reducing the apparent mechanical contact, and by eliminating non-conducting support between the joints. In other words, with my invention there is no support between a high point 75 of the second order and an oxide layer of the adjacent

In order that the annular ridge 15 and the recess 14 may be coined by a simple stamping operation, as explained below, I prefer to have each end of the bus bar 10. provided with coaxial male and female members on opposite surfaces as shown in FIG. 2. Thus, when the end of the bus bar 10 is completed, it is adapted to mate with either or both male and female portions of an adjacent bus bar.

Such a mating is illustrated in FIG. 3 wherein a nut 17 is tightened on a bolt 18 to compress the ridge 15 in the recess 14. It should be noted that the engagement of the ridge in the recess is the only engagement between the overlapping portions of the bus bars 10α and 10Z&gt; and the height of the ridge 15 is. great enough to provide a space between the bus bars at the overlapping portion of the joint. Thus, the only surface stressed by the bolt 18 is the outer curved surface 19 and/or the inner curved surface 20. In FIG. 3, the only contact is at 19. Since each integral section of the surface 19 is at an angle 0 instead of perpendicular to the bolt 18, the only engagement at surface 19 is by a wiping action which will deform the conducting metals sufficiently to rupture normal oxides and films. The angle Θ, FIGS. 3 and 4, is shown as about 35° which will provide adequate wiping action. For the best wiping action, 0 should be less than 10°. However, allowance should be made for production tolerances to insure engagement of the ridge 15 at the outer sharp corner of the groove 14. Therefore, I prefer to have the angle θ between 5° and 45° depending on the practical machining tolerances and the metal being machined, as well as the oxides of films likely to be encountered.

It should also be observed that a deeper recess 21 is provided at the bottom of the recess 14. The recess 21 is machined or punched out to a depth which will prevent any contact. A contact oyer the entire surface of the ridge 15, including the region of the recess 21, would provide a direct compression resistance to the bolt tension which would prevent sufficient strain to cause desirable wiping action at the surface 19 (or 20). Normally, I prefer to have contact at only one of the surfaces 19 or 20 to insure sufficient deformation of the metal with a reasonable bolt stress.

By limiting the contact surface, I am able to have the bolt 18 of sufficient tensile strength to compress the mating surface 19 or 20 and rupture any oxide of other film surface thereon, and to fracture any peaks between the contacting surfaces. Thus, the apparent contact or mating area, according to my invention, is actual electrical contact area.

In order to facilitate the development of the proper mechanical stresses at the surface19,1 prefer to form a convex annular ridge 15 having a slightly larger diameter than the concave annular groove or recess 14. This results in a high pressure contact or contacts formed at the outside edge surface 19 of the annular groove 14. The use of the outer ridge slightly reduces constriction resistance for a given compression because, the perimeter of the outer edge is slightly greater than that of the inner surface 20. Since the area of contact may be assumed to be proportional to the compression of the bolt, the longer perimeter provides a lower resistance. The mismatch also allows the use of multiple assemblies without severe permanent deformation of the convex ridge 15.

With a % copper bus bar 10 two inches wide, a load of 2300 pounds in the bolt 18 will rupture oxides, but will not cause permanent deformation of the. surface 19 whereby the joint may be disassembled as required and reassembled repeatedly to provide a low. resistance, connection. Since this stress is sufficient , to :rupture oxides, there need be no special cleaning of the joint either at the time of original manufacture or when reassembling the joint in the field.

Also, in FIG. 2, I have indicated a method of coining

3,002,17*3 s conductor because the stress concentration may be made sufficient to rupture all films.

Even taking into account the above factors, the theoretical resistance of the joint of my invention was several times the empirical value until I factored in the constriction resistance advantage obtained by having a line contact for a given area instead of a circular contact area. This reduced the theoretical value by 80 percent to approximate my emperical data. The fact that the line I obtain is curved to form a perimeter of a circle does not materially affect the improvement in constriction resistance.

In order to explain the relation of constriction resistance to the shape of an elliptical contact area, I have shown in FIG. 5 a curve with the constriction re- 13 sistance plotted on the axis of the ordinates, and the ratio of the major diameter of an elliptical area to the radius of a circle of the same contact area plotted on the axis of the abscissa. The actual maximum value of the constriction resistance depends on the particular metals 20 used and several other factors which have been known for some time. However, with a particular area such as one mil, the maximum constriction resistance occurs when this area is a circle. This point is plotted at the extreme left end of the curve shown. With the area 25 of one mil and similar conditions (same metals, etc.) in the form of an ellipse having a major axis five times the length of the radius of a circle of the same area, this resistance is reduced to 60 percent of the maximum value of the circle. As the ellipse is flattened to have 30 the ratio of 10:1, the constriction resistance is reduced to 40 percent and at 20:1 to 25 percent. The minimum constriction resistance factor of zero theoretically occurs at infinity when the contact is a straight line.

While I have illustrated and described particular em- 35 bodiments of my invention, other modifications will occur to those skilled in the art. For instance, the actual contact area may be more than one percent of the overlapping area of the bus bars so long as a controllable pressure may be applied, and my invention can be used 40 with other conductors than copper, with modifications depending on the hardness of the metal and hardness of the films likely to be found thereon. With aluminum, the oxide is particularly troublesome, and I have found that there must be a wiping action between the mating 45 surfaces of more than .015 to obtain the most efficient joint. I intend, therefore, to cover in the appended claims all such modifications as fall within the true spirit and scope of my invention.