[B24Spider]

Some years back I had ARP make a batch of rod bolts for Aurelias. Long conversations with one of their engineers about dry vs lubed, what type of lube, etc. (Lancia spec was always clean and dry, and they tested most of their critical fasteners, most of which were rolled threads. You can see the test marks.), and an emphasis on stretch as opposed to torque. He was explicit that new fasteners should be torqued and loosened three times before final assembly, to burnish any irregularities of the threads, which affect torque and crush. I agree.

I had a beam type originally, until my tools got stolen in '75. My first clicker torque wrench was a used Mac I found in a pawn shop for cheap. Afraid of inaccurate readings, I sent it out to be calibrated, and always wondered if it was even "out". I have and use a range from inch/#s on up to 200#s, all click type, never had any problems. My shop mate has a new electronic Snapper, but I don't like wrenches that need batteries.

 

[B24Spider]

When I started my reply last night there were 27 posts, and I didn't notice all the rest today until I had hit post (recovering from minor surgery, and staying away from sharp things for now). Perhaps that was a good thing.

Stretch is the real objective, it's measuring it that's the issue. Torque implies stretch. Also, I try to avoid running taps and dies on old fasteners if I can, as most consumer/mechanic grade ones are sized to cut for a free-running fit, and often remove more metal than I'd prefer. But, sometimes I do...

 

[AlfistiSA]

Ok folks, I'm back from a 10 day car tour with the little '57 Spider and my intro back onto the BB is a 4 page discourse on Torque wrenches with a few upsets along the way.

I've now learned more than I ever knew about torque wrenches, I'm not going to venture an opinion save I have both types and use them as and when...

For the rest, please play nicely with each other from now on - Grazie Mille

Greig

Wearing Admin hat here

 

[guillotineFI]

While I'm not so sure I really want to dip my toe in this conversation, I do have a question I've been pondering on torque wrenches lately. So, I thought I would go ahead and try to sneak this question in to the group.
Working on the 105 alfas, we are dealing with relatively low torque values for the most part - like under 100 ft-lbs, with exception of the main engine pulley (140 ft-lbs on 1750/2L). So, I've been working with torque wrenches up to 150 ft lb capacity. So, while for crank nuts maybe not the best - otherwise good.
But, lately I've been running into higher values (like on motorcycles) of 170 ft lb and a little higher being not uncommon.
So, shopping for a torque wrench for these higher values - up to 250 ft lb - in itself is simple enough.
But, here's my questions: most reasonably priced 250 ft lb torque wrenches are not much longer - if any longer - than 150 ft lb ones. Anything longer and you quickly get into rather expensive tools (they are available but pricey!).
So, what are you guys using - like for a V6 crank nut for instance (at over 170 ft lb) ? I mean, we know you can't (should not) use an extension on these wrenches right? So, do you have a husky neighbor to help yank the stubby wrench. Or, you all have 36" long $700 torque wrenches for these nuts?

 

[DPeterson3]

A hopefully brief addition to the understanding of torque, et al. It also relates to Guillotine's question about large torque values.

As a few have noted, the real question is about tension, or "pre-load", not torque. The Tension is the end result we are seeking. Torque is an inferred measurement we can use to get close to the exact amount of tension. I'm trying to be brief, so some things may be glossed over the better to make this easier to read.

I have used "Stretch bolts" for the rods in my Lycoming engine. It is a 200 HP version of another Lycoming engine that produces 180 HP. The 180 uses essentially the same bolt as the 200, but the 180HP bolts are installed using a torque wrench, whereas the 200 HP bolts are carefully measured for how much they stretch during the tensioning process.

The "science" of fasteners has shown that a fastener achieves its greatest strength and reliability at a point where its stretch is just at, or slightly into, the point where the fastener enters its "plastic" stage. The metallurgists and engineers are good at defining just how much this stretch must be. As a general guidance, stretch bolts are used only once, due to the possibility of the combination of initial tensioning and operational stresses having gone far enough. We don't typically reuse the torque-bolts either, out of an abundance of caution. However, a general practice might be to retorque a "torque bolt" if something needed to be changed during engine assembly, but not re-stretch a stretch bolt, once it has reached its target. The bolts, in 1997, were a couple of hundred dollars apiece, IIRC.

So - what is the outcome difference between the two?

The first point is the question of which is better/worse, bringing a fastener to a too-loose, or a too-tight condition?

The #1 killer of a critical fastener is being too loose, NOT too tight. With a Stretch Bolt, we try to achieve perfection, and as we are setting the tension using stretch, we are in essence measuring its tension directly, and not via an inferrence. Yeah, yeah, yeah, if the quality of that bolt and its metallurgy is all wrong, then stretching doesn't work. But, generally a $300 stretch bolt has gone through some pretty impressive design and testing.

A "torque bolt" will carry a torque setting intended to get its tension well up into a range of good and reliable tension, but without running a risk of over-tightening. We KNOW that being too loose is a killer, so if our torquing technique puts a slight extra amount of tension into the fastener, things should be OK. Undershooting the tension, on the other hand, can be terminal for that bolt, and whatever is about to be impacted by its failure. The claims of the "cam" torque wrench are probably solving a problem that doesn't really exist, ie overshooting the tension.

Dang! Page crashed. Some stuff lost...

So - a stretch bolt is a direct measurement. X stretch on Y bolts = N tension. There are no variables other than the manufacturer specifying whether to use a lubricant or not.

With a torque bolt we run into both known, and poorly defined, variables. Our challenge is to either get rid of the variables, or use techniques that minimize the impact of those variables on achieving the designer's intended TENSION. Torque is an inferred measurement, and does not guarantee that the intended tension was achieved.

More on this another day, but to the above question about BIG torque values...

Some WW2 Radial engines specified a torque method for installing those huge propellers. Use an 8 foot bar on the socket, with a 180 lb man hanging from the end.

A dirty secret is that a lot of very experienced mechanics will attach the larger crank nuts using an air impact wrench, and maybe some loctite.

 

[ossodiseppia]

Some WW2 Radial engines specified a torque method for installing those huge propellers. Use an 8 foot bar on the socket, with a 180 lb man hanging from the end.

My kind of method!

Don, I am curious why too lose is a killer. I suppose it is intuitively obvious that it's going to get effed up from vibration and any sort of movement.

Cannot "too tight" also damage a screw? If too tight, won't the threads get damaged, thus weakening them?

 

[B24Spider]

Considering those in the conversation, I thought this might be a good illustration...

Some years back a friend was ferrying a customers' Piper Cherokee from the southeast back to Nevada City, Ca. The engine was a single Lycoming opposed six.

He (with his wife and two children on board) crossed the ice covered Great Lakes, and over the Rockies with out missing a beat. As he started his final into NC, (three 90 degree turns to the right) and took the first 90, he noticed a vibration. Before he could make the second 90, there was a rather large bang, the canopy was covered in oil, and the vibration became very serious. Though it was VFR conditions, he's a pilot, not a flyer, and had been keeping an eye on his instruments, and was able to make the last 90 and touch down on a dime, for a roll to the end of the strip, where a firetruck was, fortuitously, already waiting, but the flames were already out.

Opening the cowl revealed that the middle right cylinder was halfway off the head studs. What happened? Most likely undertorqued rod bolts. The "loose" bolt gets stretched just a little bit each revolution, then relaxed, and stretched again. Eventually it fails.

A "good" slightly overtightened bolt will not stretch further under normal stresses.

-Steve

 

[Jim G]

A dirty secret is that a lot of very experienced mechanics will attach the larger crank nuts using an air impact wrench, and maybe some locktite.

.
I have absolutely no idea what your talking about. 😇

 

[guillotineFI]

Well. I suppose I have my answer then

 

[Gubi]

Well. I suppose I have my answer then

Indeed

You are correct that you can't use an extension on most torque wrenches. For stuff like the V6 nut you're referencing I'd suggest one of two things:

1) Just buy a longer wrench. For example, the Gearwrench 1/2" digital I linked to earlier goes up to 250 ft-lb, and it's about 2 feet long so it's not too hard to get to the 170 of the V6 nut if you've got a little leverage. It's also really useful for wheels and suspension components

GEARWNCH 1/2" Drive Electronic Torque Wrench 30-340 Nm -85077 - - Amazon.com

GEARWNCH 1/2" Drive Electronic Torque Wrench 30-340 Nm -85077 - - Amazon.com

www.amazon.com

2) Get a digital torque adapter. They're more awkward to use so I don't recommend them for general use, but they're length-independent. So you can stick a stupid long breaker+cheater bar on one to make it easy to get all the way to the upper limit. This one goes up to 250 ft-lb as well

ACDelco ARM602-4A 1/2” (25 to 250 ft-lbs.) Heavy Duty Digital Torque Adapter with Buzzer and LED Flash Notification – ISO 6789 Standards with Certificate of Calibration - - Amazon.com

ACDelco ARM602-4A 1/2” (25 to 250 ft-lbs.) Heavy Duty Digital Torque Adapter with Buzzer and LED Flash Notification – ISO 6789 Standards with Certificate of Calibration - - Amazon.com

www.amazon.com

I don't do this stuff for a living, so I personally wouldn't be comfortable using a bare impact wrench: too much worry in my mind of cracking the crank pulley. I've used my impact wrench many a time with torque sticks, though, and I've found them (indirectly) to have at least reasonable accuracy.

Wheels, for example. My other car takes like 110 ft-lb: I gun them on in the air using a 100 ft-lb torque stick, lower the car, and then do the final torque to 110 by hand. The bolts always turn for the final torquing. So I guess in a pinch I'd be comfortable just using the gun, an appropriate torque stick, and some Loctite on something basic like a crank pulley. But the other options above are much better ideas.

 

[B24Spider]

(NB- I am neither an engineer nor a metallurgist. This is my understanding.)

Regarding "too tight screws" (ossodisepia), elasticity (stretch), and yield of bolts and threads, metal can stretch to a degree, before it deforms, and return to its' original size. Beyond that point that it will no longer return to its' original dimension, and no longer has its' original strength. That is, it will then start to deform (yield), and eventually break.

The difference between return and deformation varies depending on alloy, condition (normalized, hardened, etc.), dimension, and service conditions, but is, for practical purposes, a (small) range with an upper bound. Damage, including rust, will lower that value. A properly designed fastener will be able to stretch sufficiently, maintaining elasticity at value, without deforming threads. Torque to Yield bolts/studs are NOT reusable.

When measuring stretch on rod bolts, their length is measured before mounting, tightened to measure (generally plus 0.00XX" of an inch), and also upon disassembly. Stretched free bolts are discarded. Not so easy to measure on lots of other fasteners (main and head studs, etc.), hence the use of calculated and experiential torque specs.

Another note on running dies on rolled threads is that a certain amount of the strength is in the continuous grain of the hardened (by the roll forming) threads, as opposed to the interrupted grain of a cut thread. Cutting these rolled threads with a die reduces strength more than cutting on "standard" cut threads.

 

[Jim G]

Another note on running dies on rolled threads is that a certain amount of the strength is in the continuous grain of the hardened (by the roll forming) threads, as opposed to the interrupted grain of a cut thread. Cutting these rolled threads with a die reduces strength more than cutting on "standard" cut threads.

Thats why you use rethreading taps and dies. They don’t cut threads.
Most of the time I can straighten the threads just by turning it by hand. Worst case. I have to put a stubby wrench on it.

 

[B24Spider]

Jim,
You're quite right. In my defense I had some minor surgery Friday; I'm still not at my sharpest, and had missed the "rethreading" amongst the sturm and drang of some the other posts. Staying away from machines.

I did leave the impression that I'm clueless; many will indeed just use a cutting tool. I don't have a general rethreading set, just a wheel stud/axle set, but will use well lubed nuts and bolts to the same effect. If it's not enough, and it's a critical fastener, it gets replaced. I should have said so.

 

[DPeterson3]

My kind of method!

Don, I am curious why too lose is a killer. I suppose it is intuitively obvious that it's going to get effed up from vibration and any sort of movement.

Cannot "too tight" also damage a screw? If too tight, won't the threads get damaged, thus weakening them?

Osso…

I’ve read the discussion on why too loose is worse than too tight. I’ve decided that’s a discussion not necessary for helping interested and willing people to embrace the correct technique for setting preload.

We’ve all twisted off a bolt or stripped threads. Therefore, too loose is better than too tight, right?

No.

My previous note was an effort to describe where are the OK upper and lower limits, and how to get there.

Essentially, a stretch bolt targets the absolute maximum strength AND reliability. A torque-to-tension targets the low to middle range of “enough tension and reliability”.

There is more “headroom” when using a manufacturer’s torque setting. There is less tolerance for under-tensioning. One does not want to exceed the manufacturer’s stretch limit.

 

[DPeterson3]

A follow up note.

I began by describing a simplified new-assembly process. The production engineer defines a process that achieves a statistically proven assembly process that achieves high reliability in the least time. This is the market that click wrenches were designed to serve.

For all intents a non-adjustable click wrench preset for “head bolts”, or “rod bolts” could be handed to the assembly technician (probably now a robot) and what results will be on the numbers.

The preset torque-break would not be a calculated, theoretical value. It will be determined by running 100 test assemblies through the line, with the results closely analyzed. When all 100 have the intended preload, the production engineer says “run the line!”

Next consideration is “use in repair and maintenance”.

Which also split into high preload-critical fasteners, and low preload.

A subject for another day. We’re packing for our flight on Thursday.

By the way…. I’ve had ten inflight failures in single-engine aircraft. Practice makes permanent.

 

[PSk]

Date: 10/26/2032
Post: I'm not sure, beam, digital or clicker torque wrench ... ?

Pete

 

[Gubi]

By then it'll be a sonic screwdriver, Pete

Look, I answered the original question back on post #7. Everything since then has just been madness and chaos.

For the record I used the GearWrench digital ones that I Iinked to all last weekend doing the rear suspension. They're pretty sweet and not super expensive, and I recommend them if anyone wants to upgrade.

 

[ossodiseppia]

Be safe, Don.

 

[DPeterson3]

Be safe, Don.

Buenos Dias,

For me, safety begins with preparation. Then, inflight decision making. Fingers crossed…

But, back to this multi-episode discussion…

We’ve determined that fasteners stretch under tension, and that engineers understand how much stretch will deliver a reliable performance.

A brief discussion of Critical fasteners vs the other ones…

A large majority of fasteners mainly just keep things attached, and are not really facing big challenges to stay alive. Cover plates, sump, cam covers, accessory attachments. For most of these, the tension levels required to stay in place are relatively low. Because the design preloads are modest, friction is not a major factor in using torque to accurately set the preload.

And this brings us to the heart of this discussion…

When using torque to set the preload several variables creep in to disrupt our hopes.

We are no longer in the world of “new assembly”. Our fasteners will have accumulated damage of various sorts. Much maintenance is performed in-situ, so there may be dirt, metal shards, and other contaminants. There could be nicks in both male and female threads, as well as the mating faces of the structure and bolt head or nut face.

For an 8mm cam cap stud/nut, with a lowish tension, if we carefully clean, inspect the threads, replace studs when required, use new nuts, and lubricate, any type of torquing process will generally give a good result. However, if all of those steps are not taken, the fastener may become “stalled” due to an imperfection. I don’t recall the stated torque at this moment, but it’s not a lot. A click type wrench may release as the fastener encounters the shard, or dinged thread, at which point you stop your tightening, and we have an under-tensioned fastener, more likely to fail than a correctly tensioned one.

I have personally encountered numerous situations like thus, but by holding the targeted torque on my non-releasing wrench, suddenly the nut moves past the tiny obstruction, and I’ll get another 90 degrees, or more, of rotation before things truly firm up.

Yeah, yeah…. A good mechanic makes sure everything is new, clean, and unblemished condition. Yeah. Right.

The above is for low-preload fasteners. The chances are that being under-tensioned may not lead to tragedy. But…. Why use a system that can allow under-tensioning to occur?

I’ll get to high-preload critical fasteners later. More complicated.

 

[Gubi]

Well, I didn't ask for Don to take the time to write a ten page summary of fastener theory. I mean, I'm no expert, but between hobbies and my career I'm reasonably familiar with the subject so I don't really need an education. Nor was I trying to tick him off.

All I asked for is a published technical reference that backs up his claim that you shouldn't use a click wrench, that you have to hold the wrench at torque to account for bolt creep, and that any of this is relevant to the 97% of fasteners on a car that aren't torque-to-yield. If this is a widely proven fact and also specified in industry then it should be pretty easy to find references stating such.

The fact that he's got time to write all this stuff but not to find a published reference is kind of telling. So in the absence of data I'm going to continue to maintain my opinion that the above info is incorrect and will just make your life harder for no good technical reason.