Milt: I think ±10% is wildly optimistic. I would guess ±30%, without taking into consideration how people use them. I think the right way to use a click wrench is to rotate at a steady speed until the click is felt, then repeat for a second click. I have no idea if this is absolutely correct, but it seems to work on TV shows, so I do it the same way.

In the end it is often good enough (IMO) to have all the bolts at 'similar torques', never mind having everything at the 'specified' torque. I'm 99% sure I could rebuild an engine without a torque wrench and have it survive just fine. A torque wrench is a useful tool, but their importance is blown out of proportion in general. How many people buy a Harbor Freight cheapy, never have it calibrated, never set it back to zero for storage, but still rely on it? Lots. And my guess is they get along fine.

Lastly, at least if you get 'click' off of every bolt you know you didn't leave one loose. Perhaps that's the primary benefit.

Agreed on that.

OK, reality check time: with the ratchet-style torque wrenches, assuming they are accurately calibrated, the mechanism will 'break' (click) when you reach the torque set. At this point you have just surpassed your desired applied torque, but the wrench will continue to apply torque beyond this point if you keep the pressure on. You can proceed to use your torque wrench as a breaker bar if you so desire! Is everyone sure thay they have not applied, say, 10% beyond that click they heard? I have a habit of repeating that click two or three times just to make sure it's consistent, but with every 'bounce' of the wrench I can't say that maybe I pushed a bit beyond.
With a beam style it's a bit different... you can hold that puppy steady at your desired torque all day. If you go a bit beyond you will see it on the scale. But on some of the beam scales I've seen, the pointer is >10% wider than each marked increment!
What Milt posted above is very shocking, that the same torque that will tension a dry bolt can snap a lubed bolt. I thought that most bolt failures occured just below the head of the bolt. If that part (the neck) is capable of a certain torque before failure should it matter whether it's in a lubed thread or perhaps just bottomed out prematurely? I can certainly see stripping the threads or pulling a stud out of a softer material , but bolt failure?

When you get really serious you don't use torque figures. One example are connecting rod bolts - there you measure bolt stretch. There are special fixtures for this.

Richard Newton

Lube makes a huge difference.
With a lubed thread much less of the applied torque is used to overcome friction. That means with the same torque you are stretching the bolt further. Imagine if you could magically remove all friction from a nut and bolt. The only force that would resist turning would be the tension/stretch of the bolt acting on the angle (inclined plane) of the thread.

As far as where the bolt tends to break...
Under the head is common. That is why there is a rounded fillet there. Where the thread starts on the shank is common. That is why really high strength bolts are often narrow and then get larger where the threads are. The third common failure point is at the threads right next to the nut. The threads inside the nut aren't distorted/stretched like the threads in the shaft are. The transition under the bolt has a high stress load and cracks can often start at the thread where it emerges from the bolt.

BTW - I like beam style torque wrenches. They don't go out of spec like the higher tech versions. They absolutely are necessary for tearing down motors when you are racing. When you are removing the important fasteners you can take note of the breakaway torques to find out if anything has started to fail. An example is if you loosen the rod bolts and one rod has less breakaway torque you need to seriously inspect that rod to find out why the bolts were stretched.

Clicker types are fast and useful too. I use one for tech inspection. You set it for 5lbs less then the wheel torque and then you can just clicky, clicky, clicky your way around the car.

If you are carefully paranoid you should be using stretch and torque. Getting the correct stretch at too low a torque value tells you that a bolt has a serious issue. It has to be extremely rare with ARP and Raceware but it has happened. With the reports of the large variety of counterfeit pieces that are now coming out of China I wouldn't be surprised to see it happen more often nowadays.

Well, well and well. So, the plot thickens, or stretches, as it were. I suspect with these latest musings that many mag case motors have been over torqued yielding the stud pulling problem. I do understand the stress on the area that contributes.

But the real question, as brought up by PZ, is how do I measure the length of the head stud while its buried in the heads and clys? Must be a "cool tool." :D

To add to the discussion:

The breakaway torque on a static, non-moving fastener is higher than the true torque applied to the fastener. Thus, for a more accurate reading, the fastener must be moving (rotating).

Sherwood

only on a Porsche BB......

I'll bet on the Corvette BB they are comparing tattoos or something like that

of course, there is also the friction difference between plated fasteners and the ones that have a black (oxide?) coating on them...

or... "Tastes Great!" "Less Filling!"

I'm sure there are some good wrenches over there. I have to say that in my very early years I assembled more than one motor (not a lot) w/o a torque wrench. I only knew enough about the sequence of tightening and I'm not sure where I got that.

These were a couple of SBC's and some Type I VW's and mostly top ends. I depended on the feel of a long breaker bar. I do think a click type of torque wrench does not give one the "feel" of getting right to that point where you just know you're going to pull that stud or break that bolt. Probably a good experiment is to clamp up some bolts and break them on purpose. And some of the nuts like to tell you how tight they are with their own voice. ;) Exhaust nuts come to mind.

The other way around.
To avoid metal-on-metal crunch, torque given is for "Wet".
Unless specified "Dry", torque values are "Wet" meaning not just the threads but the mating-surface of the bolt/nut.
Examples are the head studs: Lube on the threads and between the washers and barrel nuts.
Of course, with any torque there are stretch-factors on the fasteners and these are allowed for with the specified torque.
In the case of rod bolts, one could apply the correct torque AND measure the stretch but why bother?

Loc-Tite thread-fastener qualifies as "Wet".

OK, dry vs. wet (!)

preliminarily - if the diff. is only 10% then that is not a huge amt. as I and milt pointed out above.

Here is what others have said on this bbs, and on the old Email version of Rennlist (way back in the Holcene period):

Torque specs given in the Tech Ref books are Dry.

Use » 2/3 to 3/4, no higher, of the dry value for lubricated threads. The Factory always uses "dry" torque specs. The numbers are for clean, dry threads.


- Some have said that ALL of the torques specified by Porsche assume anti-seize, but that was a while back and I can't corroborate that.
-- Bruce Anderson

From: "Randy Hubbard" <raceware@wwnet.net> -- Dec. 1999
In general all automotive tightening torques are based on using 30W petroleum (NOT synthetic) motor oil as a lube UNLESS stated otherwise. There are also times when a moly based grease is used as a lube. This is common in suspension parts that are torqued to very high tension. There are also rare occasions when a fastener is intended to be installed "dry" without a lubricant. Many times this is to "activate" a thread locking adhesive applied to the threads or to place the bolt into yield.

We use a 'spot' of engine oil on studs and any fastener used in dissimilar metals.

I use anti-seize on some brake stuff like sliding calipers, spark plug threads, and other places where there is heat and pressure together with fasteners that must be accessible.

Mr. Hubbard was regarded as an expert, as of course, Bruce A. is. Unattributed quotes were from people I did not recognize, so take those with an additional grain of salt.

Finally, it seems ot me that the black oxide fasteners usually come oiled (part of the manf. process).

But plated fasteners don't seem to be oiled. Of course, the plating itself is quite smooth....


Now, where is Jim Sims in all this??

that's a very helpful summary. I wonder if moisture from the atmosphere makes loctite cure, otherwise maybe it is not an anaerobic process. I don't know why this little factoid has been on my mind the last day or so, haha.

Interesting about the 30w. I seem to run into statements about using a light oil. I never thought of 30w as light although I've used it many times. I also used the book torque values even with "wet." Guess I overtightened a few things. Only time I actually clean the threads and holes before assembly is when I intend to use Loc Tite. I use brake parts cleaner. I don't believe I've ever assembled anything bone dry and squeaky clean other than hardware grade nuts and bolts on toys, appliances, gates, body parts (not suspension or mounts), etc. Most of that stuff is plated anyway. All mechanical work gets a lube. Just the blown dry by compressed air, after a solvent bath is lubed to me. There is residual oil.

this is OT regarding this thread, but brake fluid is an awesome tool to break free a frozen nut. I used some old DOT 4 on some oil lines after PB didn't work, and guess what? Instant results.

I assemble the CV joint bolts bone dry and squeaky clean with brake fluid. I also clean out the female threads on the flange with Q-tips & brake fluid, but figure some tiny amt. of grease may still be present.

Most of the major auto manufacturers assemble engines using torque angle instead of torque. The quantity of stretch is easy to calculate because the pitch of the thread gives you a fixed stretch for a given quantity of rotation.

It gives you an accurate stretch but much more importantly it is very fast and repeatable. For example if you snug up the bolts and then an additional 360° rotation gives you the correct stretch then assembly is now assembly line style. If it doesn't build the requisite torque then you know you have a problem.