Choosing the Right Fastener – Part ll

By Carl Tamm

In the first part of this article, a brief mention was made regarding “conductivity” of fasteners.  This second section will address that issue.

Peculiarities of electrical connectors give rise to further thought of fasteners.  It is not exceptionally difficult, using reasonable workmanship and materials, to make an electrical connection of reasonable conductivity initially – however, to make one which will provide exceptional duration for many decades is another story.

While there are many types of “bolted connectors” used in the utility industry, for simplicity references this article will relate to flat “pad-to-pad” type connections.  The physics involved in the atomic structure of the electrical interface are too involved to address in this article, however sufficient evidence is readily available to show that a typical 4-bolt pad-to-pad connection will achieve less than 2% of the “available” surface area that will actually make an electrical connection.  Some of the earlier references to this phenomenon of area immediately surrounding the bolt holes being the only actual contact area date back over 55 years.

Kaiser Aluminum – Electrical Bus Conductors, 1957

Accepting this long stated peculiarity to be true, one might ask, “What could be done to improve that value?”  The industry has been striving to improve the area of contact “between” the pads – but the obvious option of allowing the bolts to serve as conductive paths is often overlooked.

Use of conductive washers and fasteners can provide additional current paths from the backside of the respective pads, through the fasteners!  How obvious is that?  Why do so many disregard this simplistic improvement?  Some other misunderstandings, such as reusing fasteners, will be addressed in a future part of this article.

Perhaps a simplistic understanding of fastener conductivity will help.  One may look at either conductivity or resistivity.  I prefer the former, as the numbers are more simplistic and easy to understand.  The common basis used for conductivity values is the International Annealed Copper Standard, abbreviated IACS.  The following chart indicates values for a number of materials commonly used in our industry, and a few more for reference.

Most are surprised to learn that Gold is not the most conductive material, but is significantly superseded by Silver! The attributes of gold, and its place in Contact Physics is not the subject of this article. Of equal surprise to most, is that while we all recognize iron and steel are conductive, its value of conductivity is only 2.5% IACS. This value is close enough for argument, not to get into the specific values of different alloys, and represents all common ferrous alloys, and includes “stainless” alloys.

Alloys used for connectors commonly range from 16% to about 44%. While Silicon Bronze bolts, or the common aluminum alloys used for bolts are not nearly as conductive as the materials used for conductors, they commonly do approach the conductivity of many connectors, of approximately 16% to 24%. The lesser degree of conductivity of these fasteners is due to the alloying elements used to provide the strength needed for fasteners. Still, these conductive fasteners provide 8 to 12 times the conductivity of steel.

What is the result of using steel or stainless steel bolts in connectors? While very low in conductivity, they are yet conductive, and the low value of conductivity is the reason for the “high resistance” associated with steel. The very definition of resistance is associated with the thermal rise of the material as the result of passing electrical energy through it. The physical effect is the expansion of the material due to the thermal rise. With bolts, the higher the temperature, the less clamping force provided!

To counteract this, properly designed and applied Belleville washers are used to maintain the clamping force, as Mr. Goch covered in Part 1 of this series. As Part 2 began, our purpose is to provide a bolted joint that will maintain the electrical integrity of the connection over many decades. Maintaining the clamping force is paramount to achieving this goal. The following illustration depicts the effects of the difference in the coefficient of thermal expansion between differing materials used in electrical connections.

As the illustration depicts, differing materials expand at differing rates over a given temperature rise.  This property is given stated values for different materials, known as the coefficient of thermal expansion related to the respective materials.  Both aluminum and copper alloys expand at a greater rate than steel.  The use of steel bolts, without the benefit of properly sized and applied Belleville washers will result in rapid creep of softer material of the connectors, and the joint will loosen over time.  Of course, as it loosens, the resistance will rise, and with that given rise in resistance, the thermal rise for a given current will increase.

There exist several design features incorporated in ClampStar, the result of which provide a superior connection to other types of connections, including compression connectors.  Thus another reason ClampStar provides the properly engineered Belleville washers in its fastening system!  As Mr. Goch stated in the previous part 1, the torque nuts, designed by Classic Connectors, Inc., prevent over-tightening of the fasteners, and thereby prevent over compression of the Belleville washers.

Additional information on Belleville washers will be incorporated in another series!

Click here to view and download this article in a print friendly, PDF format

4 Comments

  1. David Eccles |

    Dear Mr. Tamm,

    Maybe you could add mention of the Heat Equation H = IxIxRxt

    Where I is Feeder current
    R is contact resistance
    T is time

    Then you could mention Contact resistance measures for each case that readers might be able to micrometer test check for ..

    It would add some practical self-checking science to the case study on conductive washers advantages.

    David Eccles | Manager – Overhead Design | Overhead Design | AUSGRID

    Level BLOCK B, EAST, 145 Newcastle Road Wallsend NSW 2287 AUSTRALIA

    Reply
  2. CR Tamm |

    Hello David,

    I appreciate the comment, most especially because the article has sparked an interest! It is also obvious, by the nature of your comment, that this is not the first time you have given serious thought to electrical connections! The electrical connector interface is a rather fickle thing, and after studying them for over 26 years, I find the more I learn, the more I realize I don’t know! When one finally gets down to the quantum world of particle physics, which after all, is the root of everything, one begins to get a glimpse of just how dynamic the electrical interface is, which I often say is analogous to a tectonic fault line during a quake!

    Resistive values may be calculated and plotted, but due to the dynamic changes which occur, they are different at any given moment/temperature, and the mechanics of different connector designs do not behave the same. Thus, as a general rule, I tend to speak in rather general terms. Considering the bolted pad, (I believe they are commonly referred to as “palms” down under ), as being a moderately repeatable connection, a “good target resistance” for the typical 4-bolt pad, measured from the approximate center of one pad to the opposite side of the other, is 5 micro-ohms, measured with a “Ductor” micro-ohm meter, passing 100 Amps across the pads.

    One can spend a great deal of time, calculating resistive values of aluminum fasteners, including aluminum washers, in comparison to steel (or stainless steel) hardware, and as a general rule will arrive at an overall “calculated” improvement of about 20% (or from our 5 micro-ohm measurement, reduce it to about 4 micro-ohms. In practice however, most often one still winds up with 5 µΩ measurement either way, because initially one achieves a bit more compression with the steel hardware (nominal torque on ½” steel is 40 lbf/ft whereas with aluminum fasteners (lubricated with wax) the torque value is 25 lbf/ft. Excuse my Americanism – I am used to thinking in “foot-pounds” and you may use Newton-meters. If so, the values are 54.23 and 33.9 respectively. I am confident, that if one had the funding and time and lab space, to make up enough connections (perhaps 100 – to give a solid average), being as careful as reasonably possible to make them repeatedly the same, and cycled them enough, those properly made with steel hardware will deteriorate earlier than those with aluminum. That is a speculative statement, as I have not had opportunity to conduct such an experiment under scientific conditions. But over the years of reviewing literally hundreds of failed connectors, I have witnessed many (apparently properly assembled) connections having steel bolts that failed in service, and never one that had aluminum hardware, with the exception of those improperly assembled, over-torque, wrong inhibitor, no inhibitor. I wish I had kept a journal, shame on me.

    On the positive side, almost all of the connections I have analyzed made with steel hardware had assembly problems. Most pertinent, to categorize them in priority order of most common problem (1) no Belleville washers, (2) wrong inhibitor, (3) over torque, (4) torque reading (if one was taken) was incorrect due to seized threads – very prominent with stainless steel bolts and nuts used in combination. Those that failed, or were discovered hot prior to a catastrophic failure, if properly assembled with Belleville washers, appropriate inhibitor, etc., were very old, 40+ years.

    The real difference comes in, not in the initial established resistance, but where it is after several hundred cycles. Given that material resistive values increase with temperature, those of steel increasing at a rate of about 30% more than with aluminum, the first assumption one might venture is that the resistance across the pads would increase with temperature. Interestingly enough, those connections with aluminum fasteners will increase in resistance ever so slightly, but not enough in 100°C rise to measure, but the contact interface resistance of those with steel hardware will initially decrease! The reason is the 2.2x coefficient of thermal expansion of the aluminum compared to steel, will result in an increase in mechanical force at the interface, thereby lowering the resistance. The problem comes in with cyclic loading, where overtime, the aluminum begins to creep under the stress, and the clamping force begins to decrease. Thus is the basis of the earlier “speculative” statement about those connectors made with steel fasteners deteriorating at a more accelerated pace.

    To wrap it all up, I avoid the thermal equations due to the dynamics of the electrical interface, and the results at operating temperatures will be different than those at ambient. There are simply too many variables to keep in check. At the end of the day, my purpose is to provide longevity to connectors, which may not necessarily be attained from the absolutely lowest resistance connection to attained, which one might measure at the beginning. Thus, I strive to provide technically sound and practical reasons why I make certain recommendations. For instance, because less than 95% of the available surface between the pads does not make electrical contact anyway, the most effective way to reduce the resistance of a 4-bolt pad, is to drill a hole in the center and add a 5th bolt!

    G’day!

    Carl

    Reply
  3. David Eccles |

    Carl,

    Wow, what can I say but yes and thanks .

    I like your adding extra bolt in the four bolt connection comment (Others have seen that need and added bolts before, I think).

    I was hoping you had more Carbon Fibre Washers by design to squeeze up the metal connector surface interfaces for more electrical contact area?

    We really only use the Heat Equation in training rookies on basic concepts too.

    Empirical pragmatism then applies, although Infra Red inspection helps sometimes for fault finding. UV corona cameras can spot open metal connector fitting cracks too.

    Even looked at colour dyes for mild overheating visual. UV and aerial pollution gets that, even if someone to look timely on site.

    (Can be useful for a HV Live Line crew live working for 8 hours to changeover a pole mounted switch with temporary bridging bypass and connection points)

    We are trialing Distributed Temperature Sensing, and that on aerial conductor winds and ambient temperature variance is challenging complex.

    A few lessons from HOT-C conductors /operating temperatures of 250 Deg C plus suggests modest metal mass increases shaped for added heat sinks without too much Corona loss shaping edges can help disproportionately well.

    Linemen, ergonomic weight, human lift, and handling issues limit conductor connections and tools required to construct connections.

    If aluminum on aluminum is better than steel bolted connections, would same apply for newer alloy metals like INVAR Steel-nickel and aluminum-nickel alloys? Bolts in Steel-Nickel might have more elastic stretch range for any designed conductor operating temperature, giving less cyclic creep /stretching than steel bolts?

    Titanium alloy steel bolts?

    Regards,

    David Eccles | Manager – Overhead Design | Overhead Design | AUSGRID

    Level BLOCK B, EAST, 145 Newcastle Road Wallsend NSW 2287 AUSTRALIA

    Reply
  4. CR Tamm |

    Interesting concepts you are throwing about there! Carbon fiber washers!

    There are some metallic washers offered in Europe, the concepts of which I appreciate, but not carbon fiber. They are metallic having a series of concentric rings about the bolt hole. This was a concept I offered Hubbell many years ago, but they were too focused on taking portions of pennies out of their cost, not improving performance.

    My focus would be to eliminate washers altogether. A flanged head bolt combined with a flanged nut would be superior. Why? The focus is to eliminate interfaces. With washers, you have additional interfaces. It is not the material that is the problem, it is the electrical interface between the two metals. A little twisting of the bolt during tightening, and the natural rotation of the nut will abrade the surface underneath, which, aside from welding, produces the greatest proportion of actual contact area in the interface. A thermal fusion weld will never fail electrically – at least not from interface degradation – as that is the only method of eliminating the interface. Other methods, such as implosion DO NOT create welded interfaces, regardless of what they may tell you – it is easy to bust that myth!

    On the subject of various alloy bolts, it is typically not the bolt that stretches from cyclic loading, it is the aluminum of the connector that creeps out from under the bolt. I use an analogy when I teach my connector physics classes – which I call “Kunnekter Fizzix” as it is presented in an elementary fashion, intended to teach both the apprentice lineman and the Engineering PhD – as placing a 1 pound steel ball on top of a pound of butter, and ask them to describe the phenomenon they anticipate will occur. Inevitably, the reply is that the ball will “sink into the butter.” Then, I explain, what really happens is the butter moves out from under the ball!

    The aluminum will only support a limited compression, based on its temper and the resulting compression modulus. Using stronger bolts makes a better connection on day one, but it degrades much more rapidly with the cyclic thermal changes. It is actually the deflection of the portions of the pads outside the areas of the fasteners that result in them being separated electrically that takes us back to that illustration in the article from Kaiser Aluminum, 1957 – and that depiction of area is many orders of magnitude greater than the real area! That area would be represented by 4 circles surrounding the bolts, 9/16” ID with an annular section of only 0.030”!

    Few people recognize the true attributes of current transfer, a subject of which I am quite fond. Yes, my wife thinks I am eccentric! But it is from those attributes that I was able to design the ClampStar such that the current density is significantly less, again by several orders of magnitude, dispersed over a much greater area than that of even a compression connector! But that is a subject for another day!

    Best Regards,

    Carl

    Reply

So, what do you think ?