Title
A vertical extension, e.g. one you might use to access a bolt recessed in a channel, does not change the reading.
A “horizontal” extension, like one you might use to reach a bolt in a tight space where the head of the torque wrench can’t reach, changes the reading because the extra length of the extension itself increases the torque exerted on the bolt compared to what the wrench registers.
It’s all about leverage. Pick up a 20 lb weight with your arm fully extended and see how if feels at your shoulder. Next bend your elbow 90° and pick up the same weight. The pressure on your shoulder shouldn’t be nearly as great.
I assume SHARKBITEATTACK means a “vertical extension” in Absolute’s parlance.
From Pushing Back the Darkness
The author was someone whose opinion some people respected. I can’t vouch but he sounds to me like he knew what he was talking about (ellipses notwithstanding.)
Put in simpler terms: Because the extension will “twist” a little bit and the force that is “twisting” the extension isn’t being applied to the nut. 100 ft. lbs of total torque being applied to the torque wrench only imparts say 85 ft. lbs to the nut, because 15 ft. lbs was “wasted” twisting the extension. Total WAG, not real numbers, just for illustration.
This is why torque sticks work on air-impact guns. They absorb energy by twisting a little bit, depending on the grade of the metal.
At least…that’s what I’ve understood up to this point.
I don’t buy it. Torque sticks do absorb some of the energy (probably more accurate to say that they store the energy), but they work because impact wrenches work in short pulses instead of a sustained way.
Of course, Absolute is right that a horizontal extension increases the lever arm and thus the torque. But any extension that doesn’t change the arm length shouldn’t have any effect on the torque for an ordinary wrench. The springiness only affects the rate at which torque increases as you turn the wrench. The torque reading will still always be correct.
Yeah, that explanation may be correct for an impact wrench, but for an ordinary torque wrench that is applying a static torque, I don’t think it makes a difference.
Springs absorb energy (more correctly, as Dr. Strangelove points out, they store it). So, for an impact wrench producing a sequence of short pulses with finite energy in each pulse, the energy that goes into twisting of a vertical extension will reduce the actual force/torque that gets applied to a bolt with each pulse.
However, springs do not absorb a static force. If you are standing on a table, and then put a spring under each of your feet, the springs will compress a bit, but you will still be exerting a force equal to your entire weight on the floor.
You will expend more energy applying a given torque using a vertical extension, since some of your energy will go into twisting the vertical extension, but ultimately the torque applied to the bolt under the extension will be the same as the readout on the wrench.
I used to set up torque wrenches in a factory, and extensions do cause torque loss. If you have a short extension, made of hardened steel, the loss will be pretty small. But if you have a very long extension, made of mild steel, it could get pretty signifigant. Most torque specs are minimums, but yoy don’t want to exceed them by too much. Any energy that is absorbed by the extension reduces the clamp load that you are trying to achieve. The Machinists Handbook has all the formulas, but I am too lazy to dig them out for you guys…
But - as others have noted - energy and force are not the same thing.
With an extension, you’ll definitely have to exert the desired force through a greater distance, which implies the expenditure of more energy. But, once a stable force (torque) is achieved, what is the mechanism by which an extension modifies this?
Assuming you mean the Machinery’s Handbook, I have a copy. I’ll try to look up tonight what you’re referring to.
So I’m visualizing this in my brain, and all I can come up with, is pinching a smallish spring by the end, and twisting the spring while the opposite end is permanently fixed and not able to rotate. The spring will not perfectly transfer the rotational force you apply to it to the opposite end, because it compresses and absorbs (or stores) that force, however, it doesn’t change the fact that you applied that force - and that you must maintain a bit of pressure on it to keep it twisted.
Say you have a nut on a bolt. And you attach a spring to that nut, and a torque wrench to the spring. Rotating the wrench will compress the spring for a bit before any rotational movement happens on the nut. When the force grows enough to actually turn the nut, that energy will be expended in a non-linear, abrupt way. Depending on the nature of the spring of course, a weak spring will deflect a lot but a very rigid spring will more immediately transfer the rotational motion.
An extension on your wrench is just like this spring, but is instead an extremely rigid version of it. It will twist to some degree, depending on the quality of the metal. The more I think of it, I’m betting that the inaccuracy of the torque wrench readout is less due to the infinitesimal amount of energy that is not perfectly transferred to the nut, and is more due to the fact that the extension is all too often held at some imperfect angle.
A crooked extension will apply that rotational torque at an angle - some of it being spent inadvertently trying to “bend” the bolt. I visualize the rotational path of the nut as it spins around the bolt like an orbit. Then I visualize the rotational path of the wrench as it spins the nut as a concentric orbit. If the two orbits aren’t perfectly aligned, then the energy being put into the wrench’s orbit isn’t going to transfer directly to the bolt, but will instead be directed into whatever the bolt is “bolted” to. This will be super easy to visualize if you stick a bolt and nut into a very thin piece of metal - the metal will deflect and bend if you’re cranking on that nut crooked.
So now that I’ve buried a perfectly good point in too many words - does this make any sense to anyone?
I dug up my Machinery’s Handbook (27th ed), but I couldn’t find anything about extensions. It has a bunch of stuff about how torque is really just a crude proxy for pre-load tension, and some tables/formulas for correcting for friction in the head/threads, but that was about it. Do you have a cite for a specific section?
That’s because axial extensions don’t affect the torque delivered by a manual/static torque wrench. Torque sticks matter for impact wrenches because they limit the transmission of impact loads from the wrench to the nut; for a linear analogy, imagine driving a spike into a log using a sledgehammer - but put a big spring on top of the spike to cushion the blows from the hammer. For a manual torque wrench applying a steady/static load, a simple statics analysis will show that when you apply a torque to one end of an axial extension (via the wrench), there must be an equal and opposite torque applied to the other end (by the nut), else the extension will experience angular acceleration (i.e. it will start spinning faster and faster). the axial extension does not change the geometry of the situation at all: your hand is still applying its force F to the wrench handle at the same distance L from the nut’s axis of rotation, so the torque delivered to the nut is still F * L.
For a crow’s foot extension, it becomes possible to change the geometry. If the the extension’s axial offset between its square socket and the nut axis is E, AND that offset is aligned parallel to the torque wrench’s handle (as shown in the diagram here), then when you apply a force F to the wrench handle, the torque wrench thinks you’re applying a torque of F * L, when really you’re applying a torque of F * (L+E).
If the offset of the crow’s foot is positioned perpendicular to the wrench handle, then you haven’t changed the moment-arm of the wrench on the nut (it’s still just L), and so the wrench will be accurate.
I think the only possible issue here is if the torque wrench isn’t in a steady/static state when you reach the specified torque.
I think if you’re using your hands, it’s impossible to apply an impact to the wrench handle that is “dynamic” in the context of the torsional elasticity of a steel extension; the flesh and skin of your hands has elasticity that is orders of magnitude greater than that of the extension.
My WAG is that if the vertical extension were perfectly aligned (parallel to the bolt and dead center all the way) it wouldn’t have any effect. However, in reality, it won’t be, so you’ll be adding some lever arm between the torque meter and the nut. Since you’re adding length to the lever arm, you’re imparting more force than the meter reads.
There is indeed a tendency for long extensions to tip if you don’t brace them. In theory, the tipping should be confined to a plane that is parallel to the force you’re applying to the wrench handle, and so it should not change the moment arm.
If the extension did tip in a direction that mattered by an amount that mattered, it would probably slip off of the nut anyway. Long extensions generally require bracing to be prevent that from happening: you push on the wrench handle with one hand, and you pull on the ratchet end with your other hand. The forces of your two hands cancel each other’s tipping moment, and you deliver a pure torque to the nut without any tendency for the extension to tip.
Key phrase:
Using an extension on the drive end, however,* (except when that extension is at 90 degrees to the torque wrench frame)** will change the required torque reading on the torque wrench.*
Right – the tilt angle could be any direction, but would most likely be in the direction of the force (at right angles) and therefore have no effect.
I bet the OP’s claim is actually based on horizontal extensions like the crow’s foot, but may have been garbled along the line somewhere. I’m confident that Mike West, quoted by Z4westy is dead wrong.