Castle Nuts

[John Vardanian]

Torque settings and castle nuts... when assembling a piece with more than one fastener, one tries to tighten all fasteners equally; i.e., the same torque set. What do you do when the nut blocks the cotter pin hole? Thanks.

I see why Ferrari switched to Nylock nuts.

john

[kare]

Change the washer, or sand it down until you get it right (there really is no option).

[treue]

John,

My sources say to torque the nut to the proper amount then just enough further to line up the hole and slot. They also say castellated nuts are for low torque applications. My thought is to stop at 90 or 95%, check it then tighten just to alignment.

The best torque is a compromise between where the fastener does not loosen but also does not permanently stretch or fail under all realistic situations. Published torques are like highway speed limits, presumably based on a number of factors, but frequently these factors do not exist or have little to do with reality and always attempt to be one-size-fits-all. I think that, often, the guy that wrote the torque value guessed on the right number based on a chart he found in Machinery’s Handbook which, in turn, may be questionable. The only effective way to determine a useful torque value is to use many, many samples tightened to various values then subjected to the full range of operating situations which the joint in question will experience. The correct value would then be between where the fastener spontaneously loosened under usage and where the bolt permanently stretched or simply failed. Unfortunately, no torque chart I’ve seen notes that the values listed therein were a total guess or were thoroughly tested. This theory does not apply to the BMW E46 M3 stiffening pan (directly under the engine) bolts: torque to 43 ft-lbs (59 Nm), then turn 90 deg + 30 deg further regardless of the torque required. I could feel the bolts stretching, but they said it was OK.


Tom

[John Vardanian]

Very interesting, Tom. Thanks. I am inclined not to over tighten. I suppose, theoretically, the cotter pin will prevent the nut from backing out to the point where the nut loosens and comes undone.

I wonder if Ferrari went thru the trouble of matching nuts to bolts so they could achieve both proper torque and perfect hole alignment.

john

[kare]

When tightening a nut/bolt you aim to get a given clamping force. This is pretty difficult to measure when working, so we use something else that correlates: torque. There is no absolute correlation between torque and clamping force, friction (value of which is largely random, coming from surface roughness, hardness, accuracy in geometry, eventual deformation/dirt... etc.) puts it off and it only get worse when the torque increases. If something is wrong with the thread, it sort of "eats away" part of the torque reducing clamping force at given value of torque.

Two work-arounds are commonly used:

1) Head studs can be grabbed with a jack and pulled out to given tension. Such solutions are easy to recognize as head studs need to be very long so the stud can be grabbed with a jack (with a nut already in place). Nuts are then tightened lightly and jack released. As the nut is not under a serious load when tightened, the friction does not get to mess it all up. This is used with big parts in conditions where failures are extremely expensive (do I need to say marine engines?).

2) Using basic torque plus certain amount of degrees. Torque will get pieces together and provide basic fastening in a window where no problems occur. The degrees provide an exact amount of elastic elongation into the bolt, which correlates to an exact amount of clamping force so friction does not get to mess it all up. Gives very accurate results as it eliminates both friction and the human factor, which is why many critical parts in modern engines are assembled with this routine.

Castellated nuts are often used in low-torque applications, but this is not always the case. Who ever wrote that to Wikipedia, didn't necessary think about all the solutions that were built in the past (about 100 years)! I think it is particularly important to notice that with large bolt diameters (& high climb) over-tightening to align with next valley (anything up to 60 degrees!) can result in a serious overload. This is also why castelled nuts often have a raised bottom so they can be sanded down multiple times, before they need to be replaced.

It is also interesting that many old school master mechanics prefer working with a solid wrench instead of a torque key. This is because they want to feel that everything is "right". Working with bare hands helps to eliminate many problems that would go unnoticed using other methods. Hands are delicate instruments and with some training the accuracy can be amazing!

More than you really wanted to know (?)

Best wishes, Kare

[Jimmyr]

Kare, to support your input on hand torque, a Ferrari factory story was relayed to me by Pete Coltron. Back in the 50's a well know midwest Ferrari collector was touring the engine assembly area and noted the assembler was torquing the bolts with a simple wrench. He commented to the forman that this could be very inaccurate. The forman replied that is the way we have always done it and it works very well. Then Mr. P. A. Sturtevant stated that his company make very accurate torque wrenches and he would like to prove that this method is more accurate. A test was set up pitting the engine assembler using a simple wrench and the newest Sturtevant torque wrench. Several head bolts were torqued by the engine man using a wrench, and several more were done by Mr. Sturtevant. when rechecked by the torque wrench both torque readings were exactly to spec. The shop forman said that his assembly men always do it this way and no need to change. Several months later Pete was told that Mr. Sturtevant had shipped 8 new torque wrenches to the Ferrari factory and they were now in daily use on the factory floor. Jim

[John Vardanian]

Kare, thanks, that was a nice mini course on strenght of materials. I tend to use the torque wrench in places where the piece is fastened by more than one fastener, mostly for balance than correct torque spec. But the one thing that I took home from your post is that castle nuts are not to be over tightened just so one can achieve hole alignment. Thanks.

john