Socket failure torques

[Achilleus]

Hello all - I'm building up the equipment for a study of the torque strengths of different brands of non-impact sockets. I've looked on YT for something but all I could find was either stuff about ratchet wrenches, or some idiotic videos that didn't actually succeed in the test. I didn't find anything very useful on plain old sockets.

Does anyone remember seeing anything in this regard?

 

[russ455]

Probably Need to do your own testing

 

[2ndGearRubber]

You're going to have some issues going that. In my experience the sockets fail over time, not at a given torque. More often than not you will be rounding over normal fasteners. Increasing torque again, using a driven fastener which is taller and can grab more of the interior of the socket, you then run into snapping drive tools.

IMO it's not straight forward to test. Most stuff "fails" the max torque application test by rounding off your normal type of hardware.

 

[Achilleus]

IMO it's not straight forward to test. Most stuff "fails" the max torque application test by rounding off your normal type of hardware.

I've been looking for the strongest possible hexagonal bars. Maraging steel would be best but I can't find anything available in that shape. I would try the MP35N bolts for sports cars - at about 260kpis tensile strength, but they are $200 a pop. What I could find to be just about the strongest thing I can get hold of easily is the snapon hex bits from their allen sockets - they have the 3-inch long replacement bits for around $20. I'll cut this down and mount it into a deep impact socket with the longest possible broach depth.

 

[AEAdam]

I've been looking for the strongest possible hexagonal bars. Maraging steel would be best but I can't find anything available in that shape. I would try the MP35N bolts for sports cars - at about 260kpis tensile strength, but they are $200 a pop. What I could find to be just about the strongest thing I can get hold of easily is the snapon hex bits from their allen sockets - they have the 3-inch long replacement bits for around $20. I'll cut this down and mount it into a deep impact socket with the longest possible broach depth.

This is a waste of your time. What you will find is that almost all sockets are stronger than the bolts they are designed to turn. Testing a socket to failure won’t be easy. Literally every other part of your test rig will fail first.

More, the measure of a socket isn’t whether it breaks when you use it. It’s whether it removes the fastener without ruining the head. My guess is, your premise is wrong. None of us seeks the worlds strongest socket. The strongest sockets will be the thickest walled tool steel versions. We don’t want those.

 

[AEAdam]

More: youtube is now clogged with greedy fast talking men making videos that review products they fundamentally don’t understand. After you’ve spent 5 or 10 years turning wrenches, tell us what you think.

I used to think unboxing videos were the worst byproducts of our increasingly mindless world. The phony product reviews are worse.

 

[WildBill]

I've been wrenching for 40+ years and have broken only a handful of sockets. The few times it has happened it was either something like a dollar store level cheap chrome socket that was being used in an unsafe for its quality manner or an old decent one that had cracked. I have had them wear out to the point where I tossed them because they were rounding fasteners many, many times though. So I agree with AEAdam that testing for the breaking torque doesn't seem really useful as it just doesn't happen very often, and almost any socket will be stronger than whatever fastener you are sticking it on. Unless you are just testing total garbage store brand sockets, in which case it also seems like a waste of time because nobody who cares about socket quality enough to watch a video about it would buy them in the first place. But I don't understand some videos that get tons of likes so maybe if you do it in an entertaining enough manner it will be worth it for the magic interweb points, so if that's your goal it might be worth doing.

 

[Achilleus]

I've been wrenching for 40+ years and have broken only a handful of sockets. The few times it has happened it was either something like a dollar store level cheap chrome socket that was being used in an unsafe for its quality manner or an old decent one that had cracked. I have had them wear out to the point where I tossed them because they were rounding fasteners many, many times though. So I agree with AEAdam that testing for the breaking torque doesn't seem really useful as it just doesn't happen very often, and almost any socket will be stronger than whatever fastener you are sticking it on. Unless you are just testing total garbage store brand sockets, in which case it also seems like a waste of time because nobody who cares about socket quality enough to watch a video about it would buy them in the first place. But I don't understand some videos that get tons of likes so maybe if you do it in an entertaining enough manner it will be worth it for the magic interweb points, so if that's your goal it might be worth doing.

I am doing this as a scientific experiment. I am actually a collector of nuts and bolts of various different materials and grades, and also a collector of tools. To me this is an exciting piece of personal research to which I can apply some time. To your point - I have absolute zero interest in any social media ****.

I am doing two experiments right now - one for how resistant sockets are to rounding over bolts (on a relatively softer material), and one to test their inherent strength again fracture (on the toughest material I can find).

 

[WildBill]

I am doing this as a scientific experiment. I am actually a collector of nuts and bolts of various different materials and grades, and also a collector of tools. To me this is an exciting piece of personal research to which I can apply some time. To your point - I have absolute zero interest in any social media ****.

I am doing two experiments right now - one for how resistant sockets are to rounding over bolts (on a relatively softer material), and one to test their inherent strength again fracture (on the toughest material I can find).

I apologize, somehow I got the impression it was something for a YT channel and it didn't seem to make sense. For your own interests that's cool. Be careful testing, I had a socket break two car windows when it grenaded on me. Entirely my fault for sticking a cheap socket on an impact wrench. It would be interested to see the summery of your findings from a metallurgy standpoint.

 

[Achilleus]

Be careful testing, I had a socket break two car windows when it grenaded on me. Entirely my fault for sticking a cheap socket on an impact wrench.

I had a couple of cheap sockets (like 10mm or thereabouts) break on me many years ago but I remember the pain in my hand very clearly. I think the piece that came off one of them actually went into space, or burnt up in the atmosphere or something.

 

[rust in the eye]

I smell another useless YouTube "test" by an uncredentialed tester using dubious methods.

 

[Retired dozer fixer]

I am doing this as a scientific experiment. I am actually a collector of nuts and bolts of various different materials and grades, and also a collector of tools. To me this is an exciting piece of personal research to which I can apply some time. To your point - I have absolute zero interest in any social media ****.

I am doing two experiments right now - one for how resistant sockets are to rounding over bolts (on a relatively softer material), and one to test their inherent strength again fracture (on the toughest material I can find).

Are you really that bored?? Like everyone else has said good quality sockets are more likely to wear out than break. I also agree with most of you that bolts will fail before the tools will

 

[Pinne]

Are you really that bored?? Like everyone else has said good quality sockets are more likely to wear out than break. I also agree with most of you that bolts will fail before the tools will

And short of maybe some small / thin wall 1/4 drive stuff... the square drive should fail before the socket does anyways.

I'm sure one socket will be stronger than another, but the strength of a socket has never been the issue I've run into. It's always fit and tolerance.

 

[larry_g]

I'm sure one socket will be stronger than another, but the strength of a socket has never been the issue I've run into. It's always fit and tolerance.

This is a big consideration. What tolerance are you using for the hex you intend to use? The wedging action of the hex changes a bit by how well it fits in the socket. Do you know the min/max hex size for per SAE standards for fasteners and the standard tolerances for wrenches/sockets.

lg
no neat sig line

 

[AEAdam]

This is a big consideration. What tolerance are you using for the hex you intend to use? The wedging action of the hex changes a bit by how well it fits in the socket. Do you know the min/max hex size for per SAE standards for fasteners and the standard tolerances for wrenches/sockets.

lg

Well said.

Some people look at sockets and think “this is an extremely simple tool that anyone anywhere can design and manufacture”.

I think some of us realize sockets are unbelievable complex little doo-dads. The flats aren’t flat and the size of the opening isn’t just the max hex head size.

One of the best GJ threads on sockets was that chap who put Dykem on a bolt head and torqued with different sockets to see the location and size/shape of the contact patch. If one wanted to compare socket quality and predict socket performance in specific circumstances, that was a good way to do that.

The stress the socket sees is highly dependent on the location of that contact patch. Too near the edge, you run the risk of yielding the bolt head. Too near the center of the flat, you increase the out of plane (radial) stress on the socket and will have to make a thicker wall.

Sockets are a delicate balancing act where tiny fractions of an inch matter.

 

[Ohio Andy]

Are you really that bored?? Like everyone else has said good quality sockets are more likely to wear out than break. I also agree with most of you that bolts will fail before the tools will

Have you ever met a material scientist or structural engineer? You'd be shocked at what some of us do some fun... Well maybe even horrified.

For example, what is yourP aul Erdős number?

One time, just for fun, I calculated how long a beam that was 1 mi plus 1 ft long would deflect if bent to fit into a space 1 mi long. At this point I don't even remember what incorrect assumptions I made as to how the beam would deflect, probably assumed either a circular Arc or a quadratic....

Whatever our assumptions were, there was no close form solution. We had to simply determine numeric solution and move on.

And yeah that was just for fun, we weren't paid to do it

 

[zendriver]

Wlould any of this prove anything?

One socket "A" does this, one socket "B" does that. A company makes a million of a particular socket. The ultimate performance of each socket is probably all over the place, depending on many factors, over a given period of time.

To conclude that "A" is better than "B" (or vice-versa) would not seem to be fair without extensive, longer term testing of both.

Maybe that's the plan.

 

[Ohio Andy]

Wlould any of this prove anything?

One socket "A" does this, one socket "B" does that. A company makes a million of a particular socket. The ultimate performance of each socket is probably all over the place, depending on many factors, over a given period of time.

To conclude that "A" is better than "B" (or vice-versa) would not seem to be fair without extensive, longer term testing of both.

Maybe that's the plan.

He's saying it's for fun.

From a QA perspective, you run the math and then you know how many parts you need to test to get your interval based on the statistical probability that things are fine.

Yeah worked QA for an aerospace company. I understand what you're saying, but what he's looking for is obvious discrepancies.

Almost no one provides confidence intervals while posting testing on say YouTube. And I really don't expect someone doing something for fun to do a confidence interval.

Last time I did that well let's just say it was related to critical infrastructure