Torque specs are created by the design engineer with
new parts (underlined because it's related to the surprise ending.) Yes, any kind of lubrication will affect the bolted joint. A bolted joint is much more complicated than first glance. Essentially, you are stretching the bolt to develop enough force to hold the joint together. The stretch, measured in thousandths of an inch, comes from tightening the nut on the bolt (for a nut and bolt) or just tightening the screw (for a screw into a threaded hole). If you understand how this stretch happens, you'll better understand why lubrication affects the torque required to get the same amount of stretch (and force, since they are proportional.)
The threads on the end of the bolt or screw are in the form of a helix, just like a coil spring (keep that comparison in mind for later). The external threads on the bolt slide inside the internal threads on the nut, so you have a very narrow and long rectangular plane sliding against another, wound in a helix. Now, back to the coil spring. If you think of those planes sliding on each other in the shape of the wire of the spring, you can "unwind" them in your mind. What you end up with are two planes at an angle, and as the bolt is tightened, you are sliding one of them up the other - like pushing a heavy box up a ramp. The angle of the ramp is the helix angle. That angle can be determined by "unwinding" one complete thread to get a right triangle representing a ramp (WARNING: Trigonometry Required!) Here's a sketch I just did to show it, since I've typed a thousand words now, so a picture is in order
:
Hopefully, that all makes sense because all that gobbly-gook is necessary to understand the influence of lubrication (I think). Think of pushing a heavy box up a ramp. What's creating the resistance? Friction, caused by the weight of the box. The friction is affected by the interaction between the bottom of the box and the ramp surface. It's not germane to this explanation, but the angle of the ramp affects it, too. In the case at hand, though, the ramp angle is set by the threads and doesn't change. The weight of the box also affects it, and that is germane. In this analogy, the weight of the box increases as you push it up the ramp because the weight of the box is actually you stretching the bolt by tightening it. The more you tighten, the more it stretches, and the more the box weighs.
So, you're pushing an increasingly heavy box up a ramp. Lucky you! What could you do to make it easier to push it up? Put it on a wheeled cart! Why does that make it so much easier? Because you've drastically affected the frictional force resisting you - in a good way. Wheel bearings have very little friction in them!
Now, I've typed all this out and made a goofy sketch to get to this simple conclusion:
Putting lubrication on the threads is the same as putting your box on a cart. So now use that to think about the Kroil on the threads. With that on the threads, you'll be able to push the box more easily - meaning, it'll take less torque to develop the same stretch on the bolt. If the Kroil works fabulously, and you torque it to spec, the threads could strip or the bolt could fail in tension.
Now, I alluded to a surprise ending in the introductory sentence. Here it is: since the torque spec is determined using new parts, it's not really applicable if your parts are rusty during assembly (assuming that was the reason for the Kroil). Rust in a bolted joint can seriously affect its performance. If the internal and external threads are rusty, the friction between them can be drastically different. If the internal diameter of the sleeve is reduced by rust build-up, and the external diameter of the bolt is increased by rust build-up, the interference created will work against stretching the bolt. Even increased friction between the head of the screw or nut (whichever is turning when you tighten) can affect the actual bolt stretch.
This surprise ending means that how the lubrication affects the bolt stretch is incredibly complicated on old vehicles and very hard to determine. Using a reduction percentage is wasted effort because it's not even close to uniform for that joint ,on that vehicle, across the multitude of situations that can occur over years of driving it.
I grew up in the rust belt working on rusty stuff (thank goodness I'm not there anymore). I took two different approaches for this situation, depending on my financial status (poor vs. not poor):
- sab poor - develop a feel for a fastener yielding while being tightened (yielding means you're using so much force, something is permanently stretching prior to breaking). The yielding occurs usually in the threads, but it can also occur across the full diameter of the bolt. The nice thing about yielding is that when you transition from elastic stretching (non-permanent) to inelastic stretching (yielding), the torque at the wrench decreases, and you can develop a feel for that, just as it starts. If you stop immediately, you've just "torqued to yield" which is as tight as you can get it without failing. If the joint was properly designed (
), the bolt is not in danger of yielding further under load because the joint is now resisting the forces applied to it, rather than the bolt. The trouble with this method is that developing that feel takes many, many bolted joints and over-tightening-past-yield occurrences.
- sab not poor - use all new parts in the bolted joint and torque to spec (including whether or not lubrication was in the spec). The trouble with this method is that, in the rust belt (actually, I still do it even though I'm no longer in the rust belt), I use anti-seize on the threads and bolt whenever re-assembling. Unfortunately, when you use anti-seize, you're back to the "how much do I adjust my torque due to the lubrication?" question. My personal journey went from poor to not-poor, so by the time I could afford new parts, I already had developed the feel for yielding. If you haven't, and you're worried about torquing the bolt properly, follow the specs - only use anti-seize if it's specified. You may have to deal with rust later, but that really doesn't happen all that frequently.
I've typed another very long-winded response equivalent to the much more concise, "it depends..."
