In the March, 1991 issue of the Lotus club newsletter "ReMarque" the following letter-to-the-editor exchange between a reader (first guy) and one of the tech editors (second guy). The original article on Europa front suspensions is of no interest to us here, but the discussion on bolts, proper torquing and dry vs. lubricated fasteners is quite educational-Jack Haggerty - Tech editor ARA (Alfa Romeo Association of Northern California).
As noted in one of Dan Miller's comments on Bob Metz's article on the Europa front suspension, I also highly recommend antiseize compound. Ever since I had to buy new bodies and bearing holders for a set of Stromberg carburetors because of the galling of the aluminum thread, I've used it exclusively on almost all threaded items and slip fits and have not had another problem for 20 years. It definitely enhances future disassembly.
However, since torque settings generally assume dry threads, the settings must be adjusted lower whenever antiseize compound, oil, etc., is used. My rule of thumb is 65% of the maximum when using antiseize compound. This may sound scary, so let me explain torque settings.
The purpose of tightening a bolt is to stress the fastener under tension. Generally speaking, the tension should be set and the fastener sized accordingly to exceed the separation force between the two parts. This tension not only keeps the parts together, it also keeps the fastener from unthreading and from going through negative tension (compression), which could result in fatigue, i.e., the premature failure of the fastener before reaching ultimate (breaking) strength.
The way this tension is measured is by measuring the torque required to turn the fastener. However, this torque reading is a combination of two forces: the tension being applied acting on the angled threads, and the sliding friction between threads. As tension increases, so does the friction to slide the threads across each other. Torque settings, therefore, are calculated using the tension required in the fastener (what we want) and the friction which must be overcome to turn (slide) the threads past each other. Settings are calculated, unless specified otherwise, using dry, metal-to-metal (thread-to-thread) friction calculations. Lubricating the threads with oil, WD40 (Use Amsoil synthetic MP - it well exceeds WD40), antiseize compound, etc., reduces this friction and, therefore, reduces the torque required to obtain the same tension setting (which is what we want, but cannot directly measure cost effectively). This is the reason for decreasing the torque setting when using these materials.
Trust me. When I first started using antiseize compound, I exceeded the structure strength of several fasteners before I came across the above information Even worse is exceeding the yield strength of a fastener, which is less than its ultimate (breaking) strength, but which causes weakening and permanent deformation (stretching) of the fastener, and which could then fail later when slightly more pressure is applied during use, as with Lotus suspension failures and accidents.
This problem occurs more often when using smaller sized fasteners, which have less tolerance for error, or when reducing safety factors to lighten weight, e.g., with racing, aircraft and space vehicles.
--Dan Wysocki Stone Mountain, GA
Bob Metz's original article alluded to the fact that lubricating the threads of the suspension pin should be avoided, and this point is correct for two reasons. First, a lubricant will reduce the effectiveness of the Nyloc nut by reducing the friction that causes the nylon insert to "grip" the threads. Second, the same reduction of friction will create a great deal more tension in the bolt. This additional tension, as Wysocki points out, may exceed the fastener's elastic limit, causing it to fail and become permanently deformed.
Confusion arises when one assumes that he torque applied to a bolt will always create the same tension in the bolt. Not true! The purpose of tightening the nut is to create a specific tension in the bold, but since this tension is not easily measured, manufacturers have supplied us with a conversion of tension to torque values (which are easily measured) for a particular fastener under particular conditions.
I checked Carroll Smith's Nuts, Bolts, Fasteners and Plumbing Handbook to confirm Wysocki's point. The answer was found in chapter three, "Thread Physics." The text states that many variables affect the amount of torque required to produce a given tension, including the type of plating, the cleanliness of the threads, the fit of the threads, and the type of lubricant used (if any), as well as the thread pitch and the bolt size. Further, it states that, in dry, unplated threads, about 50% of the torque is used to overcome friction between the bearing surfaces and the work faces (where the nut meets the washer or part), and about 40% is used to overcome thread frictions A quick calculation shows that only 10% of the torque is used to create tension in the bolt. It follows that reducing this friction (by applying a lubricant) will cause a much greater portion of the same torque to be applied to creating tension.
Alternatively, the tension created by applying identical torque to a lubricated bolt thread (and bearing surface, also) will be much greater than that produced in a dry thread. How much greater? A graph from the Smith text (pp. 48-52) shows that about 200% more tension - TWICE THE TENSION! - is created when using antiseize, and close to 150% more tension is created when using engine oil as a lubricant. It should be noted that torque tables for particular applications are quoted as "dry and clean," unless they specifically state otherwise. So, if you choose to lubricate the threads and bearing surfaces of a fastener application, the "dry" table no longer applies; instead, refer to a table that shows lubricated torque values or reduce the torque used to produce the recommended tension.
Because of the number of variables involved, I'm not sure that Daniel Wysocki's rule of thumb (65% torque when lubed vs. dry) will apply in every case, particularly with smaller fasteners, where the tolerance of error is smaller, but his guideline is more appropriate than using the "dry" table on a lubricated bolt.
Since the only accurate methods for measuring tension (special stress sensitive washers, stress indicating bolts, and actual measurement of bolt stretch) are usually impractical in automotive applications, I prefer to use the known table and accompanying method (dry vs. lubed) for critical fasteners.
For those interested in the more technical details of the theory and practice of using fasteners, I would highly recommend reading Carroll Smith's Nuts, Bolts, Fasteners and Plumbing Handbook. This is the latest book from the author, who is an expert on constructing and maintaining race cars that finish races. The old adage, "to finish first, first you have to finish," applies here. In his first book, Prepare to Win, Smith spent a couple dozen pages on the theory and practical applications of fasteners - basic nuts-and-bolts stuff. This latest book goes into much greater detail, providing the reader with an in-depth look at the chemistry of metals, the engineering terms of forces exerted on fasteners, and thread physics.
All of this may sound like a dry textbook, but engineer Smith explains the theories and background info in language that is understandable to the layman and avoids a lot of engineering jargon. Another of Smith's books, Engineer to Win, goes into even more detailed chemistry, metallurgy and technical background, but this book emphasizes the practical application of theory and shows diagrams and photos of both good and poor examples. It also includes a useful source directory, so that you can find "the right stuff" and avoid the myths that Smith refers to as "moonglow."
--Dan Miller From the Lotus Club newsletter, "ReMarque"