[July 17, 2012]

When, and how, to use Belleville washers [Electrical Apparatus]

(Electrical Apparatus Via Acquire Media NewsEdge) Sorting through the contradictory advice for determining the right tightening torque (ProQuest: ... denotes formulae omitted.) "CLEAN, DRY, AND TIGHT" - ELECTRICAL circuit maintenance in a nutshell. Tightness of terminal screws, to eliminate the high resistance "glowing connection" phenomenon so often resulting in a destructive fire, has been widely publicized for many years. Some authorities have argued unsuccessfully that the National Electrical Code should include specific tightening torque for all connections.

Yet torque alone may not be a complete solution to the problem. At least as controversial an issue has been retention of the initial clamping force exerted by properly tightening fasteners at installation. Periodic re-tightening has been advocated, and often practiced. However, re-tightening can make matters worse by distorting mating parts or screw threads.

A popular solution for connections between flat surfaces such as busbars is addition to the joint of controlled spring compression, to retain initial clamping pressure despite dimensional changes within the joined members. This is the function of cone-shaped Belleville washers, known also as "disc springs" or "conical compression washers" (Figure 1 ). Patented by Frenchman Julien Belleville in 1867, they've been around since long before the electric power industry, and are widely used in automotive, aircraft, piping, and other mechanical assembly applications.

"Spring" in the joint is desirable under two conditions. One is extreme thermal cycling. Repeated heating and cooling of the joint, involving different thermal expansion rates of conductors and fasteners, causes dimensional variations that lead to relaxation of initial joint compression. Such changes are not always fully reversible.

A second condition is the inherent behavior of aluminum, widely used for industrial bus conductors. Under sustained pressure, aluminum is subject to cold flow or "creep," in which the metal shrinks in the direction of the applied pressure. Proper use of Belleville washers sustains the desired clamping pressure despite conductor creep. Such usage is therefore common practice with aluminum; less often for copper-to-copper joints.

What else causes loosening of a bolted connection? No matter how smooth a flat surface may appear, microscopic surface irregularities are always present. Clamping pressure squeezes those variations into intimate contact, minimizing electrical resistance across the interface. Obviously, inadequate pressure - too low a bolt tension - means poor contact and high joint resistance. If moisture can creep into the loosened joint, surface corrosion develops, resulting in still higher resistance. But too much bolt tension can be equally troublesome, because excessive tightening of clamping bolts cam damage threads, further compromising joint integrity.

Joint loosening may involve several other phenomena. They include: * Embedment relaxation. Those microscopic high spots forced into intimate contact can be so highly stressed that local yielding occurs. This leads to relaxation of bolt tension. Up to 10% of the initial joint preload may eventually be lost.

* Vibration. Preload is lost slowly at first, then more rapidly as mating surfaces of fastener threads begin to slip past each other.

* Elastic interactions. In a joint containing multiple bolts, all cannot be tightened simultaneously. Various patterns are followed to tighten them sequentially at successively higher torques. But as a second bolt is tightened, increased joint compression causes the first bolt to relax somewhat, and so on throughout the process.

The advantage of the Belleville over a conventional splitring lockwasher is in applying clamping pressure along a 360° circle rather than concentrating force at a single point.

In the simplest electrical connection involving a single through-bolt, one Belleville washer is placed under the nut (less often, beneath the bolt head), which is then drawn up enough to flatten the washer. Once the Belleville is squeezed from its original slightly coned shape into complete flatness, tightening torque abruptly increases, a point readily apparent to the installer "by feel." No torque wrench is needed. The washer design assures that the tension in the bolt will then have a specific value. For example, one manufacturer offers standard washers for %-inch bolts with flattening loads ranging from 2,000 to 4,300 lbs.

In Figure 2, the relationship between that "flattening load" and the corresponding washer deflection appears as the plot (a). (It's not actually a straight line, but a curve shaped according to washer dimensions, as shown in Figure 3.) For such a seemingly simple device, without moving parts and embodying no operating "process," the design of the Belleville washer is surprisingly complex. Here, for example, is the formula for the spring load F in terms of the dimensions in Figure 4: in which: ...

F = spring force, lbs.

E = modulus of elasticity; for steel, (30)( 10<>) lb/in.

µ = Poisson's ratio; 0.3 for steel K1 = a constant dependent upon the ratio Dq/Di, typically about 0.7 ? = washer deflection under load P D0, Dj, h, t are all dimensions in Figure 4 When the washer is flattened so that D = h, this becomes: ...

Calculation of stress within the spring is even more complex. Fortunately, choosing and using Bellevilles properly doesn't require familiarity with such details. But keep in mind that, as one consultant has said, "Belleville washers are not toolbox items to be used indiscriminately ... do not use such washers without the guidance of your firm's engineering force or that of the washer manufacturer." Multiple Belleville washers may be combined in many ways (see Figure 5). In the "parallel" arrangement of Figure 2(b), the deflection is the same as for a single washer but requires twice the flattening force. In the "series" combination of Figure 2(c), the deflection is doubled for the same force. More complex series-parallel combinations, such as Figure 2(d) or Figure 5, are seldom encountered in electrical work.

Probably the most widely debated question concerning a bolted joint using Belleville washers is this: Once flattened, should the washer be left that way? A closely related question is: How should the flattening force be related to the limit on tensile stress within the bolt? Electrical engineers, washer manufacturers, and industry standards have provided varying answers to both questions. Most authorities do agree that the bolt should initially be tightened until the washer is flat. At that point the abrupt increase in tightening torque then signals that no further effort is needed.

But what then? Here are examples of the recommendations to be found in the literature throughout the past 40 years: A. Leave the joint alone, with the washer flattened * Electric utility engineering department: "Tighten until Belleville is flat." * Washer manufacturer: "It is not necessary to 'back off' the nut after tightening." * Electrical connector manufacturer: "It is best to just flatten the Belleville. Experience and testing have shown this procedure to be satisfactory." * Washer manufacturer: ". . . many plant procedures call for the technician to tighten the bolt until the Belleville becomes flat, and then 'back off' 1A turn . . . this practice is unnecessary [because] after a single thermal cycle, the spring unflattens slightly anyway." * Connector industry consultant: "Theoretically, a Belleville should be flattened and then the load slacked off slightly so there is room for expansion. However, this can be a tricky maneuver. . . . Tightening until the washer is flat is acceptable." * Aluminum industry consultant: "Some writers have advised backing off the nut after flattening. . . . This is not . . . recommended by the Aluminum Association, wire and cable manufacturers, or the manufacturers of Belleville washers. The industry recommends tightening until the Belleville is flat, and leaving it that way. The relaxation of the metals under stress will be sufficient, within 48 hours, to restore the crown to the washer, and it will retain its full compressive force on the connection. . . . These recommendations are based on both laboratory tests and field experience." On the other hand . . .

B. Do not leave the washer flattened * Consulting electrical engineer: "The proper technique is to draw the washer to the flat position . . . then the nut is loosened approximately one-twelfth turn. This provides for thermal effects which would tend to tighten the connection beyond the design point." * Electric utility standard: "The joint is torqued until the Belleville washer is completely flattened The nut should then be backed off Veth of a turn." * Switchgear manufacturer: "You shouldn' i flatten the Belleville. You should compress it to 80% of its free height to give you room for expansion and contraction. Based on our experience, properly designed and torqued Belleville washer joints will not work loose in the field." * Technical editor: ". . . the manufacturer may say to flatten the washer and then back off slightly. Make sure you don't back off too far, because when you flatten the washer a second time [why would you do that?], it will have less clamping force than it was designed to have." * Electrical conductor handbook: 'Tighten nut until Belleville lies flat. . . . Then slack off ... so there is ample margin for maximum thermal expansion." In typical examples, the desired nut loosening was calculated at from about V4oto ?? of a turn, depending upon the design of the washer.

* Electric utility engineer: "Manufacturers of Belleville washers recommend maximum deflection limited to 75% to avoid sharply increasing force and stress characteristics. This is easily measured with a feeler gauge after assembly." * Trade magazine article: "Conical spring washers must not be left at the fully flat position. . . ." * Electrical consultant: "Since a Belleville is a spring . . . optimum performance is obtained by not completely flattening, and allowing for movement in both directions. . . ." Choosing the right number of washers How many Bellevilles per bolt? When joining only two relatively thin conductors, a single washer on one side of the joint is customary. The clamping pressure applied at the washer rim will carry through to the single interface between the two conductors.

At least one switchgear manufacturer uses only a single Belleville even for joints of three or four aluminum bars. Other users are concerned that when multiple conductors are to be joined, contact pressure will tend to be dispersed at successive interfaces away from the washer, so that two washers (the "series" arrangement of Figure 2(c)) are desirable, as shown in Figure 6(a). One authority favors that arrangement any time the total joint thickness reaches 1 Vi inches.

The relationship between tightening torque and bolt preload is imprecise. Eliminating the need to decide on that relationship is therefore a considerable advantage in joint assembly. The basic formula is well-known. For unlubricated threads, the force in pounds = T/(kd), in which T is torque in lb-inches, d is nominal bolt diameter in inches, and k is a so-called "friction factor" usually taken as 0.2 for steel bolts. Surface finish makes a difference, though; plating can cause the factor to range between 0.15 and 0.33; 0.22 is suggested for galvanized hardware.

For a %-inch bolt, as an example, the required torque to achieve a 5,000 lb. clamping load will be (0.2)(5,000) (0.375), or 375 lb-in. (31 lb.ft.). But what is the recommended bolt load? That depends first upon the bolt material, and upon the thread design. Most electrical connections are made using Grade 5 steel bolts. Also widely used, however, are bolts of silicon bronze, brass, or other proprietary alloys, for which a recommended load limit must be obtained from the supplier.

For Grade 5 steel only, however, with standard coarse threads, the maximum bolt loading varies widely as these figures indicate, compiled from a number of sources: Similarly wide ranges apply to the corresponding tightening torque recommended by various sources. Those torque values also exhibit a wide range: Why so much variation? Although Grade 5 bolts are most often recommended, many published load limits do not identify the bolt grade to which they apply.

Also, one supplier has acknowledged that "the clamp load is calculated by arbitrarily assuming usable bolt strength is 75% of bolt proof load times tensile stress across area of threaded section." That assumption, sometimes as low as 40%, isn't always made clear.

Whatever the bolt load, choosing the material, size, number, and spacing of the bolts is beyond our scope here. (See "What makes bolted connections tight enough?" in EA March 1988.) A good current reference is NEMA Standard CC 1 , "Electric Power Connections for Substations." Most common bolt sizes are % and Vi inch. One basic difference of engineering opinion should be noted, though, when multiple bolts (typically four or more) are needed because of joint size. Here are two views: * Major connector manufacturer: Since "bolt force increases in considerably smaller ratio than the area of the bolt ... in order to get high contact pressure, it is desirable to use many bolts of relatively small cross-section." [Yet the most frequently published values appear to contradict the premise - in the bolt size range from 1A to Vi inch, the average bolt load limit varies in about the same proportion as the bolt area.] * Leading switchgear supplier: "To achieve high torque and large contact area, it is preferable to use a few large bolts rather than many small ones." Once that load value is selected, the next choice is the flattening force of the Belleville washer to be used. Matching that to the desired bolt load eliminates the need for tightening the bolt to a specific torque. But here again, recommendations are inconsistent. One author advises the user to "make sure the torque for your Belleville washers matches that of your bolts." That's not much help, because washers are not rated by torque but by their flattening force. What the author intended was a match between washer flattening force and bolt tension.

But how close a match? These are typical recommendations: * Connector consultant: "For example, the Belleville washer on a %-in. Grade 5 bolt should flatten out at between 4,000 and 5,000 lbs. . . . You must use a Belleville washer that flattens out above 4,000 lbs." * Electric utility engineering: "For % in. bolt, maximum flattening load is 2,000 lbs, for bolt torque of 20 lb-ft. [which corresponds to a clamp load of 3,200 lbs.]." * Consultant again: "Required flattening force should be close to that of the bolt used." [How "close"?] * And yet again: "A Belleville should flatten at a force close to, but less than, the optimum force exerted by the bolt." [What's "optimum"?] In any event, the flattening involves only a small deflection - 0.02 to 0.04 inch. Because the "dish" is so slight, Belleville washers are often found to have been installed upside down. The clamping force is improperly applied, resulting in poor joint contact. Another installation caveat: Neither the Belleville nor an associated flat washer should be larger in diameter than the width of the joined members. Otherwise, some of the pressure applied by the Belleville will be "in the air." The flat washer will bend away from the Belleville instead of fully transmitting clamping force to the joint. That force will thus be reduced to an unpredictable extent.

A Belleville washer naturally exerts spring pressure only along its outer periphery. To distribute that pressure more evenly throughout the joint, a flat washer is normally applied under the Belleville - see Figure 7. The resulting increase in contact area can be seen in Figure 8. One authority advises that, 'The flat washer should be of the same material [as the Belleville]. Stainless steels are recommended for all electrical connections to minimize eddy current heating and local electrolytic action where moisture is present." On the other hand, a switchgear manufacturer has asserted that, 'There are many types of equipment, which have been in service for 50 years, or more using carbon steel bus joint hardware." Because stainless steel hardware is not always readily available, and stainless fasteners tend to exhibit lower tensile strength, standard steel is more commonly used, and is often plated. If the Bellevilles are also plated, to deal with a hostile environment, be sure to avoid the electroplating process. In the presence of hydrogen, an electroplated spring can become brittle and disintegrate under impact or vibration. "Mechanical zinc plating" (galvanizing) is an alternative. Eddy current heating in joint assembly hardware does not appear to be a problem drawing significant attention.

Though not as simple as it looks, the carefully selected and properly used Belleville washer can be an important ingrethent in a trouble-free electrical installation. During planning, work closely with hardware suppliers, because no "universal" standard applies.

Typical high-power bolted electrical terminal connections to which Belleville washers might be applied, particularly when aluminum is involved.

As one consultant has said, 'Belleville washers are not toolbox items to be used indiscriminately ... do not use such washers without the guidance of your firm's engineering force or that of the washer manufacturer* By Richard L. Nailen, P.E., EA Engineering Editor (c) 2012 Barks Publications

[ Back To SIP Trunking Home's Homepage ]