Does torque alone loosen a nut?

[TimNJ]

Greetings,

I’m thinking about designing a DC injection brake for tools run on AC induction motors. It’s been done many times before, yes.

If you have a tool with a reverse threaded spindle (e.g. some sort of saw), there is some propensity for the arbor nut to come loose if aggressive braking is applied.

My question is mostly a question of physics: Is torque the only factor for loosening a nut? Most people are familiar with an impact wrench, which applies impulses of high torque. A related question is: Does the “impulse” part of an impact wrench actually matter(?) or is this simply a means of achieving increased torque, by storing energy and releasing as a strong blow?

I’m wondering if the deceleration of the spindle can be regulated/optimized in such a way to maximize deceleration, but without loosening the nut. Does abrupt application of braking torque loosen a nut any differently than if you ramp up to that same torque?

Thanks,
Tim

[Jeroen3]

The amount of torque  by braking will be determined by the inertia and the deceleration. It does not matter if your turn this by hand or apply impulses on it by braking abruptly. The forces are still applied.
The friction of the thread surfaces determine the breakaway torque. This friction exist because of deformation of the cut thread.
There is no other mechanism for loosening a nut, but you can increase the breakaway torque with serrated flanges, glue, lock rings or lock wire.

This is all variable though, material, tightening torque, greases, wear, temperature.

iirc deceleration and acceleration curves can be made friendly by using the S-curve, if that is available to you.

[rich t]

> Is torque the only factor for loosening a nut? Most people are familiar with an impact wrench, which applies impulses of high torque. A related question is: Does the “impulse” part of an impact wrench actually matter(?) or is this simply a means of achieving increased torque, by storing energy and releasing as a strong blow?

An impact wrench works by releasing stored energy.  The shorter the angle it does this over, the more torque it can apply.  Just like "force times distance" in a line equals energy, "torque times angular displacement" about a point also equals energy.

If you watch an impact driver work to loosen a nut, you will see initially it moves the nut only infinitesimally with each impact...  This is because the nut is tight and the torque required for movement is high, so the available energy can only move it a small angular displacement...  As the nut loosens, and the required torque for movement reduces, and the nut moves much further with each impact (of the same energy).

To answer your question, yes, it is only torque that loosens (or tightens) a nut -- you can apply it gradually steady state with no impulse and achieve the same result -- this is how a torque wrench works, approaching and never exceeding the desired torque.  (In general, torque wrenches should never be used for loosening a nut, and impact wrenches should never be used for tightening one -- despite what you see at tire repair shops!)

What is really happening as the nut is tightened is the bolt is stretching, causing a normal (or clamping) force across the face of the nut against the thing being clamped (and also on the threads), which in turn causes the friction coefficient between the nut and the thing being clamped (and also the bolt threads) to resist the nut rotating.

The equation for friction between two objects is simply that the available (sideways) friction force is the friction coefficient (typically a value between 0 and 1) times the normal force (perpendicular to the friction force).  It is this friction that the torque needs to overcome to rotate a nut.

(In some materials there can be "stiction" as well, which is a static force that needs to be overcome for initial movement, which adds to the friction, but this is small in the nut/bolt scenario.)

[rich t]

> Does abrupt application of braking torque loosen a nut any differently than if you ramp up to that same torque?

Oh, yes it can for sure if the fastener itself is heavy compared to what it is fastening...

In general, fasteners should be extremely light compared to what they are fastening so that the force of momentum in the fastener is negligible.

Fasteners that are directly on a high-speed radial load (like the nut on a drive sprocket on a motorcycle) may also have a locking mechanism like a bent washer to help secure it.

[TimNJ]

Thanks to you both.

So, if torque can be applied in bits or continuously, as to create enough rotation to overcome friction, then the “suddenness” should not matter? Are we oversimplifying reality at all?

A constant deceleration is produced by a constant torque, right? So, in the case of the S-shaped curve, you approach the torque gradually, but once you’re there, the force/torque is the same. The S-shape curve seems to improve oscillation/overshoot of the target velocity. Probably good for stepper drives and stuff. In my example, could some sort of similar resonance induce loosening?

[TimNJ]

Vibrations can reduce stiction?: https://physics.stackexchange.com/questions/419221/can-vibrations-be-used-to-reduce-friction

And vibration does not necessarily result in a net torque?

[David Hess]

Does abrupt application of braking torque loosen a nut any differently than if you ramp up to that same torque?

An abrupt application of torque is different.

When torque is applied gradually, the entire assembly is strained evenly according to the elasticity of each part.  When torque is applied suddenly through impact, two things are different; strain occurs on the parts closer to the points where torque is applied before parts which are further away, according to the speed of sound in the materials, and also because of this, the inertia holds parts in the assembly in place before strain can propagate to parts further away.

So because of those things, impact is more effective for overcoming friction than gradual force.

As far as DC braking, which is incredibly effective, ramping up the braking force will easily avoid impact which would loosen things while having little effect on braking time.

[pcprogrammer]

My question is mostly a question of physics:

Nothing real to add, but just a comic note.



How scientists address this physics problem. 

[T3sl4co1l]

Another perspective: how fast do you think you can dI/dt into the stator winding?  How much torque do you think will be transmitted from the rotor (massive) to the nut?  A fraction probably: the nut, and whatever's on the end, are modestly rigidly linked by the shaft, and this does form a CLC lowpass filter in the torsional mode, but the placement of the poles depends on the inertias (inertiae?) and spring constant, and the nut depends on the torque, effectively the current into the last C.

I suspect the answer is, overall, comparable to a hump of mains frequency, i.e. 10ms give or take, and the poles somewhat less than that, but could be more if there's a big pulley or other flywheel thing on the end, which would also be the case that maximizes torque.

If you're spinning it down in some 100s ms, I doubt anything notices.  Deceleration is inherently sigmoidal, though not along any particular function: attack is limited by dI/dt, decay by, well, the exponential decay of the braking force vs. RPM.

Tim

[TimNJ]

Does abrupt application of braking torque loosen a nut any differently than if you ramp up to that same torque?

An abrupt application of torque is different.

When torque is applied gradually, the entire assembly is strained evenly according to the elasticity of each part.  When torque is applied suddenly through impact, two things are different; strain occurs on the parts closer to the points where torque is applied before parts which are further away, according to the speed of sound in the materials, and also because of this, the inertia holds parts in the assembly in place before strain can propagate to parts further away.

So because of those things, impact is more effective for overcoming friction than gradual force.

As far as DC braking, which is incredibly effective, ramping up the braking force will easily avoid impact which would loosen things while having little effect on braking time.

Thanks. That makes pretty good sense to me, although I can't really visualize all the force vectors over that brief timeframe to fully understand what the net effect is. That said, from an "intuitive" perspective, it surely seems that "soft" de/acceleration would be beneficial in this context.

[TimNJ]

Another perspective: how fast do you think you can dI/dt into the stator winding?  How much torque do you think will be transmitted from the rotor (massive) to the nut?  A fraction probably: the nut, and whatever's on the end, are modestly rigidly linked by the shaft, and this does form a CLC lowpass filter in the torsional mode, but the placement of the poles depends on the inertias (inertiae?) and spring constant, and the nut depends on the torque, effectively the current into the last C.

I suspect the answer is, overall, comparable to a hump of mains frequency, i.e. 10ms give or take, and the poles somewhat less than that, but could be more if there's a big pulley or other flywheel thing on the end, which would also be the case that maximizes torque.

If you're spinning it down in some 100s ms, I doubt anything notices.  Deceleration is inherently sigmoidal, though not along any particular function: attack is limited by dI/dt, decay by, well, the exponential decay of the braking force vs. RPM.

Tim

Thanks! So what do you think is  "comparable to the hump of mains frequency" exactly?

And true, a "sudden" application of torque is not even that sudden, given the system has inertia and its own transient response. And, assuming most stators of 0.5-2HP motors are in the range of 10-30mH or so, then it might take ~10-30ms to get up to "full torque" (whatever that is), if you hit it with a fast rising voltage step. But I have no idea if that ramp up is fast or slow in the context of the system inertias, spring constants, etc.

[jpanhalt]

You raise an interesting question about rise time.  It is, of course a hammer effect.  The motor, whether electric or air isn't just starting and stopping, but rise times have been studied.  I don't usually get access to full articles in Elsevier, but the first reference (a review) was accessible.  The second link is to an article cited in the first.  It's just an abstract.

https://www.sciencedirect.com/science/article/pii/S016981412030651X

T. Armstrong et al.  as cited in above:
https://pubmed.ncbi.nlm.nih.gov/9973877/

[T3sl4co1l]

I am a bit curious, myself, how much it matters whether the stator is supplied from a low impedance (voltage source), or biased anywhere inbetween on the Thevenin equivalent spectrum.  I don't know offhand how much AC is generated, how soon after application of voltage or current, in an induction motor.  Point being, absorbing the reflected energy (which is now not just eddy currents, but being generated from rotation) may have a sharper impact than the general rise time.  But there's still the leakage between stator and rotor (air gap etc.) that limits peak torque.

And yeah, time constants comparable to mains (half)period, or say, starting torque but just in the opposite direction.

Overall, I doubt it makes a difference, and you might consider braking more gently just in case.  Somewhere between "keeps spinning for minutes" and "brick wall" there likely exists a happy medium of "it's finger-safe in as much time as I can lift my hand from the switch to $spinnythingy", without unlocking the nut in the process.

And if you can get some Loctite(R) in there, that's always a possibility.

Tim

[David Hess]

I know that applying even a moderate voltage causes extreme braking, and it is easy to achieve damaging levels of jerk/jolt.

[johansen]

Greetings,

I’m thinking about designing a DC injection brake for tools run on AC induction motors. It’s been done many times before, yes.

If you have a tool with a reverse threaded spindle (e.g. some sort of saw), there is some propensity for the arbor nut to come loose if aggressive braking is applied.

My question is mostly a question of physics: Is torque the only factor for loosening a nut? Most people are familiar with an impact wrench, which applies impulses of high torque. A related question is: Does the “impulse” part of an impact wrench actually matter(?) or is this simply a means of achieving increased torque, by storing energy and releasing as a strong blow?

I’m wondering if the deceleration of the spindle can be regulated/optimized in such a way to maximize deceleration, but without loosening the nut. Does abrupt application of braking torque loosen a nut any differently than if you ramp up to that same torque?

Thanks,
Tim

the induction motor won't be stiff enough to provide any more braking torque than about twice its breakdown torque, assuming you use dc injection breaking to put about 4 times the nominal full load amps through the stator.

my estimate is entirely that, i once modified a 230vac motor for 138v and plugged it into 208 three phase. it started and got up to 3600 rpm in about half the time it would have taken at 230, when wired for 230. its no load amps were 17-18 iirc, when the adjusted full load amps for 138vac would be 9.3 amps. 

anyhow, with regard to impact guns, the axial impact of the driver being driven into the fastener is significant, but not every gun produces an axial impact force.

[TimNJ]

I am a bit curious, myself, how much it matters whether the stator is supplied from a low impedance (voltage source), or biased anywhere inbetween on the Thevenin equivalent spectrum.  I don't know offhand how much AC is generated, how soon after application of voltage or current, in an induction motor.  Point being, absorbing the reflected energy (which is now not just eddy currents, but being generated from rotation) may have a sharper impact than the general rise time.  But there's still the leakage between stator and rotor (air gap etc.) that limits peak torque.

And yeah, time constants comparable to mains (half)period, or say, starting torque but just in the opposite direction.

Overall, I doubt it makes a difference, and you might consider braking more gently just in case.  Somewhere between "keeps spinning for minutes" and "brick wall" there likely exists a happy medium of "it's finger-safe in as much time as I can lift my hand from the switch to $spinnythingy", without unlocking the nut in the process.

And if you can get some Loctite(R) in there, that's always a possibility.

Tim

I was thinking about a CC/CV flyback, primarily running in CC (with an adjustable limit) and with an exaggerated soft start. CV is only there as a protection really. Add a big relay plus logic/timers/delays etc.

Then I figured a simple SCR half wave rectifier should do just as well(?). Just ramp up the on-time/phase angle for a similar soft start effect.

I don’t know what the commercial solutions do, but I’ve seen big chonky 24Vac transformers in some (but not all?).. Maybe SCR on the output of the transformer? I wonder if that’s supposed to serve as some sort of inherent power limitation, just in case. Otherwise, I figure you’d just rectify straight off mains, save yourself a transformer. I feel that direct regulation of DC stator current might be a little more appealing in some ways, but I don’t know.

Anyway, that’s mostly off-topic…