To avoid these ads, REGISTER NOW!

Tool Tolerances and Fitment

teagueo

Well-known member
Joined
Mar 7, 2017
Messages
463
With the Snap-On vs HF battle going on, it was interesting to see alloy composition and tolerances pop up in the discussion.

Tighter tolerances are more expensive to manufacture due to more careful machining and higher reject rates. And you can only go so tight before stuff stops fitting.

On the left is standard clearance and on the right is much tighter than normal:


Tolerance_Comparison_TRO.png

With less clearance, you're able to increase the area in contact with the socket. With all that force distributed over a larger area, there's less stress on the socket and also lower chances of rounding off the fastener.

On the left you have concentrated forces on a small area, vs a much wider area on the right.


Contact_Patch_Comparison.png

Obviously, the more of a fastener you have engaged, the better. The wider and tighter fitting wrench is usually going to win with less collateral damage.
 
To avoid these ads, REGISTER NOW!

Stubby1743

Well-known member
Joined
Jul 16, 2023
Messages
716
Location
UK
The desirability of tight tolerance tools is of reduced practical use if the tolerances of the fastenings are not also tight. In an ideal world we would all be using tight tolerance tolls on equally tight tolerance fasteners, but we do not live and work in an ideal world. :(
 

CHI_Tool&Die

Well-known member
Joined
Jul 20, 2021
Messages
1,379
Location
Chicago, IL
Fasteners ****. They are all over the board so I don’t need the tightest tolerances on my sockets and wrenches. It’s also why I have so many slip joint, water pump, and cobra, adjustable and plier wrenches.

Also, I don’t know if tighter tolerances are always just found on expensive wrenches. I have a bunch of Heyco SAE wrenches I bought as a set for like $70 and they are the tightest fitting wrenches I own.
 
OP
T

teagueo

Well-known member
Joined
Mar 7, 2017
Messages
463
I've never had a tool not fit onto a fastener (yet).

I think tool manufacturers could get a bit closer to the upper range of fastener sizes. A near perfect fit is awesome when it does happen.

In the case of the pics above, spark plug hexes are more tightly controlled and spark plug sockets are universally sloppy fits.
 

AEAdam

Well-known member
Joined
May 27, 2023
Messages
2,737
Location
SE PA
Let's be a little more precise. Wrench or socket clearance and tolerance are 2 different things and they seem to be getting mixed up here.

Let's start with bolts. The specs for bolts all show unilateral (negative) tolerances. A bolt with a 13mm head should never be bigger than 13mm. A bolt with a 13 mm head could measure 12.57mm to 13mm per the ASME B18.2 spec. The nominal (preferred) width across flats is 13mm. Most decent manufacturers will make bolts as close to 13mm as they can. Its possible your favorite third world country may make all bolts with undersized heads because they use less steel and are cheaper to ship.

Wrenches and sockets are all designed with a clearance to nominal. A 13mm wrench will never measure 13mm across the flats. Snap On adds approx .10mm or so. The wrench opening is desgned to have that clearance. I don't know what the tolerance or allowable variation from 13.10mm Snap On or any other manufacturer permits. I would assume, depending on the manufacturing process, factories would skew the nominal (increase the clearance) to assure no 13mm wrenches left the factory that measured less than 13mm.

Conclusion: because the variation in bolt heads is always negative, it makes sense to buy the tightest fitting tools you can get.
 
OP
T

teagueo

Well-known member
Joined
Mar 7, 2017
Messages
463
Let's be a little more precise. Wrench or socket clearance and tolerance are 2 different things and they seem to be getting mixed up here.
Agreed on the tolerance vs clearance.

What I had intended to say was; if you're going for low clearances (tool to fastener gaps), it generally requires tighter tolerance control on the tool opening. Otherwise you'll end up making too loose or overly tight of a fit.

It's also risky for tool manufacturers making tools that won't fit the wide variance in fastener heads.

But there's absolutely room for improvement. It's easy to broach a 10.15 mm hex socket that will always fit 10 mm hex heads that measures 9.9 mm across the flats. Sounds like nothing, but that's quite a bit of play and it's not getting a great purchase on the fastener except the very corners of it.

Not really an issue unless it's some irreplaceable fastener or something you really don't want buggered up.
 

308guru

Well-known member
Joined
Jun 17, 2017
Messages
459
Tighter tolerances are more expensive to manufacture due to more careful machining and higher reject rates.
Why are you assuming higher reject rates? That would just be a stupid business model. Making a lot of widgets and only keeping the "good" ones that meet a requirement is not putting food on the table.

With tighter tolerance higher precision processes and tools usually come along. I'm not putting something that belongs on a jig borer on a drill press, etc. If a process and accompanying tools are selected carefully and correctly the reject rate could be zero or close to it.
 
OP
T

teagueo

Well-known member
Joined
Mar 7, 2017
Messages
463
Why are you assuming higher reject rates? That would just be a stupid business model. Making a lot of widgets and only keeping the "good" ones that meet a requirement is not putting food on the table.
It's a lot easier to scrap parts at +.0000" / -.0005" than at +/- .005". I'm speaking generally, not from my own processes.

In a small scale shop, you can take your time and carefully creep up on dimensions, repeatedly.

Large tool companies probably aren't bringing parts to within tenths on the lathe with some scotch Brite. They don't really need to - they make tools that have acceptable fitment with the widest range of fasteners, that doesn't require insane precision.
 

RoninB4

Well-known member
Joined
Jul 22, 2020
Messages
3,498
Location
Under My House
Almost every time the manufacturing process changes plant locations, the steel supplier is different, a new QC dept. head, a new engineer, a new manufacturing engineer, or the tooling is getting worn there seems to be an "adjustment" to tolerance, finish, or fitment. Not supposed to because there are drawings with specifications on them but it happens anyway. Sometimes unusual circumstances affect products as well. I posted about this with hex keys from Eklind some time ago. Unless you're an insider at a company there's really no telling what goes on when a product becomes different and companies don't like for customers to know unless it's a selling point. JMO
 

Hakeem

Well-known member
Joined
Jan 22, 2024
Messages
1,246
Location
Chicago
Why are you assuming higher reject rates? That would just be a stupid business model. Making a lot of widgets and only keeping the "good" ones that meet a requirement is not putting food on the table.

With tighter tolerance higher precision processes and tools usually come along. I'm not putting something that belongs on a jig borer on a drill press, etc. If a process and accompanying tools are selected carefully and correctly the reject rate could be zero or close to it.
On a side note, are jig borers still used in production today? I was told they have been replaced by modern CNC machines
 

Steve_P

Well-known member
Joined
Sep 15, 2010
Messages
5,182
There are ANSI tolerances for all of this. I wouldn't expect a statistical analysis of $2 sockets to match Snap On for standard deviation of hex flat size. So.....unless HF sockets are out of tolerance.....
 

Dave455

Well-known member
Joined
Mar 19, 2013
Messages
5,801
Location
Sussex, England
Let's be a little more precise. Wrench or socket clearance and tolerance are 2 different things and they seem to be getting mixed up here.

Let's start with bolts. The specs for bolts all show unilateral (negative) tolerances. A bolt with a 13mm head should never be bigger than 13mm. A bolt with a 13 mm head could measure 12.57mm to 13mm per the ASME B18.2 spec. The nominal (preferred) width across flats is 13mm. Most decent manufacturers will make bolts as close to 13mm as they can. Its possible your favorite third world country may make all bolts with undersized heads because they use less steel and are cheaper to ship.

Wrenches and sockets are all designed with a clearance to nominal. A 13mm wrench will never measure 13mm across the flats. Snap On adds approx .10mm or so. The wrench opening is desgned to have that clearance. I don't know what the tolerance or allowable variation from 13.10mm Snap On or any other manufacturer permits. I would assume, depending on the manufacturing process, factories would skew the nominal (increase the clearance) to assure no 13mm wrenches left the factory that measured less than 13mm.
Absolutely spot on!

I’ve had this clearance / tolerance discussion with many customers in the past!

For info, the ISO Metric tolerance, for a 13mm wrench is 13.04mm to 13.24mm, so your .10mm estimate for Snap On is probably bang on too!

The better the tool, the tighter the manufacturing tolerance, so they can manufacture towards the lower end of the clearance limit. The tools are still all within the standard tolerance though, so those concerned that less clearance might mean a wrench won’t fit, shouldn’t be concerned.
 
Last edited:

BrandonV

Well-known member
Joined
Jun 9, 2023
Messages
4,030
Location
Arizona
There are ANSI tolerances for all of this.

That's what I don't get. All I see on YouTube & TikTok with this "argument" is people checking wrenches against nylon & metal bars not ANSI grade bolts & fasteners.
 

AEAdam

Well-known member
Joined
May 27, 2023
Messages
2,737
Location
SE PA
Why are you assuming higher reject rates? That would just be a stupid business model. Making a lot of widgets and only keeping the "good" ones that meet a requirement is not putting food on the table.

With tighter tolerance higher precision processes and tools usually come along. I'm not putting something that belongs on a jig borer on a drill press, etc. If a process and accompanying tools are selected carefully and correctly the reject rate could be zero or close to it.
I'm completely lost. Of course tightening tolerances increases rejection rates. How is it that a person who seemingly knows what a jig borer is doesn't know that?

Not trying to be snarky. You must have a point and I totally lost it. Try us again.
 

AEAdam

Well-known member
Joined
May 27, 2023
Messages
2,737
Location
SE PA
But there's absolutely room for improvement. It's easy to broach a 10.15 mm hex socket that will always fit 10 mm hex heads that measures 9.9 mm across the flats. Sounds like nothing, but that's quite a bit of play and it's not getting a great purchase on the fastener except the very corners of it.

Not really an issue unless it's some irreplaceable fastener or something you really don't want buggered up.
Just keep in mind: when we talk about this, people kinda assume the inside of a socket is a hex shape. Far from it! Thanks to Snap On, the "flats" inside a socket aren't typically flat at all. So as the socket rotates due to slop, the contact patch doesn't change proportionately. Flank drive isn't just corner relief.

The Flank Drive patent which defines the concept has expired, but the detailed design, the specific shape of the inner profile, remains Snap On IP. My guess is, folks have tried to copy Snap on's shape. But that's harder than one may think. My guess is, the tolerances on the broach, is in the tenths region (.0001") or maybe smaller. That's pretty tiny and not easy to measure. Its a complicated shape. .001" different from Flank Drive probably would perform noticeably differently. Snap On's clearance to nominal is only about .004". That's a little thicker than a human hair or a sheet of printer paper.
 

RoninB4

Well-known member
Joined
Jul 22, 2020
Messages
3,498
Location
Under My House
On a side note, are jig borers still used in production today? I was told they have been replaced by modern CNC machines
-That depends upon who you talk to. They're still in use (sometimes) by people that own them. Have they been replaced by CNC machinery? For the most part yes. CNC is faster and more capable of continuous complex surfaces. They also have the capability of multiple tool changes written into the program for different operations. This leads some to feel CNC is superior. Is it? For some purposes yes but not all. CNC machinery is also subject to inaccuracy from wear like any other machine. Just because it computer controlled doesn't make it more accurate. Same goes for digital vs. analog, some foolishly believe digital (or metric) is more accurate. I've not read anything that would prove that erroneous assumption. The jig borer, once called the "mother of all machines", had the purpose of locational accuracy built into the design. When used with a boring head it can produce a truly round hole. CNC often create a hole through circular interpolation, which I feel is often NOT truly round. If the tolerance/accuracy of the geometry will accept that error then it doesn't matter. Many designers aren't aware of this and a lot of times it really doesn't matter. I believe a single point cutting tool produces a superior round shape much as I prefer a jig ground hole over one made by a wire EDM. That's my opinion and my experience. Is CNC better? Like many tools it depends upon what you need to do with it.
 
Last edited:

CHI_Tool&Die

Well-known member
Joined
Jul 20, 2021
Messages
1,379
Location
Chicago, IL
I'm completely lost. Of course tightening tolerances increases rejection rates. How is it that a person who seemingly knows what a jig borer is doesn't know that?

Not trying to be snarky. You must have a point and I totally lost it. Try us again.
Tighter tolerances do not equate to higher rejection rates for a high precision shop. The whole manufacturing process would be highly refined to be as efficient as possible. I believe that is his point.
 
To avoid these ads, REGISTER NOW!

zendriver

Well-known member
Joined
Dec 10, 2014
Messages
29,810
Location
Indiana
For those that think they know how Snap On sockets are made, Here a video on how they are actually made.


Seems like the high volume process can be a bit imprecise and a lot of the final processing is making them look pretty.
 

AEAdam

Well-known member
Joined
May 27, 2023
Messages
2,737
Location
SE PA
For those that think they know how Snap On sockets are made, Here a video on how they are actually made.


Seems like the high volume process can be a bit imprecise and a lot of the final processing is making them look pretty.
Can you be more specific? What is it that looks imprecise to you?
 
OP
T

teagueo

Well-known member
Joined
Mar 7, 2017
Messages
463
Just keep in mind: when we talk about this, people kinda assume the inside of a socket is a hex shape. Far from it! Thanks to Snap On, the "flats" inside a socket aren't typically flat at all. So as the socket rotates due to slop, the contact patch doesn't change proportionately. Flank drive isn't just corner relief.

The Flank Drive patent which defines the concept has expired, but the detailed design, the specific shape of the inner profile, remains Snap On IP. My guess is, folks have tried to copy Snap on's shape. But that's harder than one may think. My guess is, the tolerances on the broach, is in the tenths region (.0001") or maybe smaller. That's pretty tiny and not easy to measure. Its a complicated shape. .001" different from Flank Drive probably would perform noticeably differently. Snap On's clearance to nominal is only about .004". That's a little thicker than a human hair or a sheet of printer paper.
Most sockets (even cheap ones) are using some type of "flank drive" geometry now. If you look at the sidewall of the hex, it's almost never flat but actually looks slightly curved.

Teague7690~2.jpg

The tighter the fit on the hex, the more length of the "flank" you're engaging. If you had dimensionally perfect bolt heads, the best you can do is have half of the hex flat length in contact, with just enough slop for rotating the socket.

image (1)~2.png
 

Dave455

Well-known member
Joined
Mar 19, 2013
Messages
5,801
Location
Sussex, England
For those that think they know how Snap On sockets are made, Here a video on how they are actually made.


Seems like the high volume process can be a bit imprecise and a lot of the final processing is making them look pretty.
It is a high volume process for sure, but that actually looks pretty precise to me.

They’re cold forging then heat treating, which is about the best way to do it currently, and every step in the process looks optimum to me.

I note they don’t actually show much of the internal forging process, but I wonder if these critical steps are commercially sensitive?

The commentary says “a series of forming tools”, and I have no doubt that is the case.
 

zendriver

Well-known member
Joined
Dec 10, 2014
Messages
29,810
Location
Indiana
Can you be more specific? What is it that looks imprecise to you?
Cut off a chunk of coiled steel rod, and then stamp it into a socket head maintaining a very tight tolerance at high volume production run?

Seems a bit dubious, but I’m not an engineer, but if it is constant close tolerance, why can’t the Asian manufactures do the same? They surely use a similar process.

:dunno:
 

308guru

Well-known member
Joined
Jun 17, 2017
Messages
459
Tighter tolerances do not equate to higher rejection rates for a high precision shop. The whole manufacturing process would be highly refined to be as efficient as possible. I believe that is his point.
Without spending time to type out a number of examples, yes, that is exactly my point.

If the process, equipment, and tooling is capable there is no reason tighter tolerances mean more scrap.
 

308guru

Well-known member
Joined
Jun 17, 2017
Messages
459
And yes, newer CNC machines seem to have generally replaced jig borers and other manual equipment. I don't visit all that many shops but either they don't use old equipment like that or they are in a corner covered in dust because they have been replaced or nobody knows how to use them.

One shop I use for some parts will mill (cnc) holes I call out as reamed with a 0.0005" total tolerance. The ability to do this relies heavily on the experience of the machinist and the capability of the machine.
 
Last edited:
OP
T

teagueo

Well-known member
Joined
Mar 7, 2017
Messages
463
Tighter tolerances are more expensive to manufacture due to more careful machining and can lead to higher reject rates if the tolerances exceed the capabilities of the production equipment.

That all better for you @308guru? Thought it'd be fairly obvious but you need incredible word precision on GJ.
 

2ndGearRubber

Well-known member
Joined
Mar 24, 2014
Messages
14,185
Location
Pittsburgh
Tighter tolerances do not equate to higher rejection rates for a high precision shop. The whole manufacturing process would be highly refined to be as efficient as possible. I believe that is his point.

I think the point is, all else equal, a tighter tolerance will involve more faults over an infinite number of widgets.

No matter what your equipment it's easier to produce something to the nearest 1/8 inch than the nearest 8 thou. This can be mitigated somewhat with QC, monitoring deviation of time, a predictive replacement prior to wear taking things outside tolerance.

There was a lot of paranoia when snap on released FDX sockets with their much tighter fit. One possibility for this innovation was nefarious, simply reusing worn out tooling. As the tooling pushes the broach into the metal, wear occurs. Eventually the tooling is under spec, and needs replaced. This is because it no longer fully "opens" the 19mm socket to 19.10 mm or whatever their spec is. This worn tooling may only produce a 19.02mm socket, which could theoretically be used as a specialty application socket. No evidence of this, but that was one theory.
 

CHI_Tool&Die

Well-known member
Joined
Jul 20, 2021
Messages
1,379
Location
Chicago, IL
And yes, newer CNC machines seem to have generally replace jig borers and other manual equipment. I don't visit all that many shops but either they don't old equipment like that or they are in a corner covered in dust because they have been replace or nobody knows how to use them.

One shop I use for some parts will mill (cnc) holes I call out as reamed with a 0.0005" total tolerance. The ability to do this relies heavily on the experience of the machinist and the capability of the machine.
As a fourth year apprentice my job was to mill dowel holes with +/-.0002 size and +/-.0004 location tolerances on a CNC VMC. It was easy as long as you had the right tooling, programs, and time to be careful. You just had to stand by the mill and check the sizing after each operation with a precision gage pin. So you can imagine it isn’t the most economical way of doing things.

The preferred way of handling holes with such tight tolerances was to run them on a jig grinder or even a WEDM. The machinist could prove out the first hole and then let it run all night unsupervised. It was cheaper, more reliable, and more efficient to run holes that way. Only real downside was that the run times were much longer and thus the productions times slowed down.
 

AEAdam

Well-known member
Joined
May 27, 2023
Messages
2,737
Location
SE PA
Cut off a chunk of coiled steel rod, and then stamp it into a socket head maintaining a very tight tolerance at high volume production run?

Seems a bit dubious, but I’m not an engineer, but if it is constant close tolerance, why can’t the Asian manufactures do the same? They surely use a similar process.

:dunno:
Yeah, if you are unfamiliar with manufacturing, I could see how that video might not look impressive. From 10,000 feet up, I would expect the basic process would be similar in factories around the world. You might even say, identical.

It's the details in manufacturing that make the difference, and they aren't obvious. The tolerances inside a socket are EXTREMELY tight. Those tolerances are a function of the broaching dies, their material, their surface finish, and the machines that drive them. The force required to cold broach steel is tremendously high. Any moving part on the broaching machine will be subject to wear. Inspecting that wear is no laughing matter. All those machine components would need constant lubrication. That lubrication needs to be clean. The holding fixtures that hold the stock must be super rigid. When you watch the video, you can see a lot of things moving, but everything looks rock solid.

Later in the video we see some workers loading fixtures but not holding anything. There are skilled workers grinding etc. Surface finish is important for high strength. It's not just there for pretty. Even the smallest scratches can become stress concentrations leading to failures later.

It has been pointed out in the past that the markings shown are for a Williams socket. It is believed both Snap On and Williams sockets are made in the same plant in Milkwaukee Wisconsin. That has caused some here to believe they are therefore identical sockets. Factories don't work that way. The broaching machines would need frequently maintenance, the broaching dies would need to be renewed. The time sockets spend in grinding, polishing, can greatly impact their performance and cost. We know from TTC tests the Williams tools are not equivalents to their Snap On cousins. My guess is, at a certain point, a machine will switch from making SO sockets to making Williams. It could have a little more wear, more slop, run broaches longer, and this is how factories can produce 2 products for different costs.

Here is a Chinese socket factory video I found on YouTube:
Notice the basic operations are similar. But testing of sockets similar to these indicates much poorer performance. So we know the output is not the same. Reasons why probably have more to do with what we don't see than what we can. Steel quality and heat treat precision would be my first thoughts. We know from testing, the steel quality varies pretty dramatically.

In general, the machines appear to be much older and lacking the automation present in the US plant. The machines, the floors all look filthy. If this was a restaurant, I wouldn't eat there. People who work in factories (as I do) care about such things.

Time =0:38 those look like some sort of punches getting manually loaded. They are using towels as guards over their heads.

T=0:54 they are using rollers to straighten the round stock before it goes into a filthy shear. Output has scrap along with product. Keep in mind, in manufacturing, the quality of any upstream operation affects the down stream operations. So even a simple shear operation can impact the finished part.

T=1:03 they are making bit sockets, holding them under a machine by hand, no safety glasses, of course. Notice the mandrel doesn't appear to be lined up with the center of press. The worker appears to be protecting her hands, keep them out from under the press. Yikes

At 1:19 notice how beat up the extension looks, huge burr around the ball bearing detent hole. Assume they haven't gotten to grinding yet, but we didn't see anything like that in the Snap On video. That mark came from a low quality machine upstream from where we are, since the ends have already been machined/formed.

At 1:50 the unfinished sockets appear stored in filthy, rusting containers in what appears to be a building with no climate control. That's going to be a problem at plating. Any rust pits would be a problem structurally. That looks like a lot of stock waiting to be plated.

2:27 packaging. Packaging often requires labor. Here is a long line of workers, cooled by shop fans, loading sockets into every one's favorite blow molded cases. Looks like a lot of cost. Imagine if those were American union laborers, each making $20/hr, with healthcare, 401k, maybe pension benefits. Modern factory could be $100/hr or more for them (each). Makes sense why we don't see those sorts of cases from US manufacturers.

This is probably a good look at what the factories look like that produce all your favorite imported tools. Our factories looked like this in the 1950s. Know that these folks WILL catch up and your continued unwavering support will help them do that. Other industries are more state of the art in China. But industry is a team sport, and they need everybody to be functioning in the 21st century or issues they are currently having will continue.
 
Last edited:

Junkdrawer Dog

Well-known member
Joined
Jan 14, 2019
Messages
1,460
Location
LV NV
The desirability of tight tolerance tools is of reduced practical use if the tolerances of the fastenings are not also tight. In an ideal world we would all be using tight tolerance tolls on equally tight tolerance fasteners, but we do not live and work in an ideal world. :(
Many inexpensive, flange headed bolts have so much draft angle on the flats from the cold heading that it can be hard to tell what size they were supposed to be. They may be 1/2 at the top and 13 mm at the bottom, near the washer face.
 

AEAdam

Well-known member
Joined
May 27, 2023
Messages
2,737
Location
SE PA
Without spending time to type out a number of examples, yes, that is exactly my point.

If the process, equipment, and tooling is capable there is no reason tighter tolerances mean more scrap.
Got it now thanks.

FWIW: Coincidentally this is a conversation I have with new engineers. We learn that in general, tighter tolerances mean higher costs through higher rejection rates. We are taught this. They use bell shaped curves to indicate variation.

My reality (after 34 yrs in industry) is that if a process (e.g. good CNC mill) produces a certain variation (e.g. +/-.005" in aluminum), changing the tolerance from +/-.030 to +/-.010 won't change the rejection rate (tho the shops typically WILL charge more for the tightened tolerance).

My other reality is that variation from a milling operation doesn't fit well under a bell curve. Variation doesn't just happen at random. There are reasons for variation, and those reasons are predictable, sometimes preventable. We can also create engineering tolerances reflective of variation to reduce rejections and increase number of acceptable parts.

Last, my personal opinion is that CNC machining has gotten SO good, engineering has become pretty sloppy at tolerancing. Our modern manufacturing processes over deliver.
 

zendriver

Well-known member
Joined
Dec 10, 2014
Messages
29,810
Location
Indiana
Yeah, if you are unfamiliar with manufacturing, I could see how that video might not look impressive. From 10,000 feet up, I would expect the basic process would be similar in factories around the world. You might even say, identical.

It's the details in manufacturing that make the difference, and they aren't obvious. The tolerances inside a socket are EXTREMELY tight. Those tolerances are a function of the broaching dies, their material, their surface finish, and the machines that drive them. The force required to cold broach steel is tremendously high. Any moving part on the broaching machine will be subject to wear. Inspecting that wear is no laughing matter. All those machine components would need constant lubrication. That lubrication needs to be clean. The holding fixtures that hold the stock must be super rigid. When you watch the video, you can see a lot of things moving, but everything looks rock solid.

Later in the video we see some workers loading fixtures but not holding anything. There are skilled workers grinding etc. Surface finish is important for high strength. It's not just there for pretty. Even the smallest scratches can become stress concentrations leading to failures later.

It has been pointed out in the past that the markings shown are for a Williams socket. It is believed both Snap On and Williams sockets are made in the same plant in Milkwaukee Wisconsin. That has caused some here to believe they are therefore identical sockets. Factories don't work that way. The broaching machines would need frequently maintenance, the broaching dies would need to be renewed. The time sockets spend in grinding, polishing, can greatly impact their performance and cost. We know from TTC tests the Williams tools are not equivalents to their Snap On cousins. My guess is, at a certain point, a machine will switch from making SO sockets to making Williams. It could have a little more wear, more slop, run broaches longer, and this is how factories can produce 2 products for different costs.

Here is a Chinese socket factory video I found on YouTube:
Notice the basic operations are similar. But testing of sockets similar to these indicates much poorer performance. So we know the output is not the same. Reasons why probably have more to do with what we don't see than what we can. Steel quality and heat treat precision would be my first thoughts. We know from testing, the steel quality varies pretty dramatically.

In general, the machines appear to be much older and lacking the automation present in the US plant. The machines, the floors all look filthy. If this was a restaurant, I wouldn't eat there. People who work in factories (as I do) care about such things.

Time =0:38 those look like some sort of punches getting manually loaded. They are using towels as guards over their heads.

T=0:54 they are using rollers to straighten the round stock before it goes into a filthy shear. Output has scrap along with product. Keep in mind, in manufacturing, the quality of any upstream operation affects the down stream operations. So even a simple shear operation can impact the finished part.

T=1:03 they are making bit sockets, holding them under a machine by hand, no safety glasses, of course. Notice the mandrel doesn't appear to be lined up with the center of press. The worker appears to be protecting her hands, keep them out from under the press. Yikes

At 1:19 notice how beat up the extension looks, huge burr around the ball bearing detent hole. Assume they haven't gotten to grinding yet, but we didn't see anything like that in the Snap On video. That mark came from a low quality machine upstream from where we are, since the ends have already been machined/formed.

At 1:50 the unfinished sockets appear stored in filthy, rusting containers in what appears to be a building with no climate control. That's going to be a problem at plating. Any rust pits would be a problem structurally. That looks like a lot of stock waiting to be plated.

2:27 packaging. Packaging often requires labor. Here is a long line of workers, cooled by shop fans, loading sockets into every one's favorite blow molded cases. Looks like a lot of cost. Imagine if those were American union laborers, each making $20/hr, with healthcare, 401k, maybe pension benefits. Modern factory could be $100/hr or more for them (each). Makes sense why we don't see those sorts of cases from US manufacturers.

This is probably a good look at what the factories look like that produce all your favorite imported tools. Our factories looked like this in the 1950s. Know that these folks WILL catch up and your continued unwavering support will help them do that. Other industries are more state of the art in China. But industry is a team sport, and they need everybody to be functioning in the 21st century or issues they are currently having will continue.
I don't mind being talked-down to, for lack of education, but seriously, you are supporting your argument, with a video of some post-Mao shithole factory, that makes socket tools for "Greatneck" and sets that might sell at Big Lots? :wtf:

China is a pretty big country that seems to be able to make a lot of really high quality stuff. No chance of any modern, high precision automated tool factories, even if they don't happen to produce a YouTube video? :dunno:

FWIW not graduated college, but have worked in production metal shops (somewhat similar to the video), where the housekeeping was really not all that. The notion that they cannot product consistent close tolerance and quality, is a silly one, IMO.
 

AEAdam

Well-known member
Joined
May 27, 2023
Messages
2,737
Location
SE PA
I don't mind being talked-down to, for lack of education, but seriously, you are supporting your argument, with a video of some post-Mao shithole factory, that makes socket tools for "Greatneck" and sets that might sell at Big Lots? :wtf:

China is a pretty big country that seems to be able to make a lot of really high quality stuff. No chance of any modern, high precision automated tool factories, even if they don't happen to produce a YouTube video? :dunno:

FWIW not graduated college, but have worked in production metal shops (somewhat similar to the video), where the housekeeping was really not all that. The notion that they cannot product consistent close tolerance and quality, is a silly one, IMO.
Didn't mean to offend you. You mentioned you didn't think much of the Snap On manufacturing video but couldn't really say why and I was just trying to help.

GJ often talks about factories but when the subject comes up, I'm always left with the impression that few here have ever really stepped foot in one, let alone spent the last 34yrs in one as I have.

Sorry the video I found didn't fit your narrative.

Adam
 

2ndGearRubber

Well-known member
Joined
Mar 24, 2014
Messages
14,185
Location
Pittsburgh
Got it now thanks.

FWIW: Coincidentally this is a conversation I have with new engineers. We learn that in general, tighter tolerances mean higher costs through higher rejection rates. We are taught this. They use bell shaped curves to indicate variation.

My reality (after 34 yrs in industry) is that if a process (e.g. good CNC mill) produces a certain variation (e.g. +/-.005" in aluminum), changing the tolerance from +/-.030 to +/-.010 won't change the rejection rate (tho the shops typically WILL charge more for the tightened tolerance).

My other reality is that variation from a milling operation doesn't fit well under a bell curve. Variation doesn't just happen at random. There are reasons for variation, and those reasons are predictable, sometimes preventable. We can also create engineering tolerances reflective of variation to reduce rejections and increase number of acceptable parts.

Last, my personal opinion is that CNC machining has gotten SO good, engineering has become pretty sloppy at tolerancing. Our modern manufacturing processes over deliver.


Let me preface this as saying my machining experience comes from using neglected brake lathes, and working in "service" and "ground floor" level jobs where I am the end "doer of things":


Is the process in your example, changing required tolerance from 30 down to 10 thou not causing higher rejection, only true if the equipment remains in a given condition and has a given output quality? But, would this not fall off after bushings/ways wear, articulating points get sloppy, etc? Does this industry simply allow this not to happen? I just don't come from a world where things are maintained. If the machine needs consumable bits or cutters, and these are run well beyond expected life, wouldn't the chatter and finish issues really throw the total tolerance into question?

In my world view, perhaps the machine can produce to 0.001 repeatably, as an example number. After improperly moving the machine several times, possibly knocking it over, doing zero maintenance or repair of worn components, improper lubricant choice, will this not drift substantially? I've heard the term "clapped out" thrown around when people are buying used machine tooling, so some amount of wear must occur. My view is the machine will produce to 0.001 for a given time, then tolerance will widen. Machinists would then be belittled and shouted at. Machine will continue to degrade, blame will continue, until firings of machinists occur. Eventually turn over will be high enough to fire whoever is actively choosing to not maintain equipment or total machine failure will occur, which will then necessitate replacement or repair, bringing tolerance back to 0.001. In my position of admitted ignorance, it seems like a skilled operator and a machine in proper condition should be able to hold a given quality. It'll just be harder to do so and you can't be as sloppy. But all that assumes the machine is kept in a way to allow this. The machinist will never be 100% perfect, so wouldn't the tighter tolerance towards the minimum the machine can produce, over an infinite number of widgets, cause more defect/reject parts simply because the error margin is lower?

Is the mentality of such production facilities that which eliminates drift and tolerance issues due to negligence? My initial thought when reading that was "yeah right, they'll run that machine with worn out joints floppier than a limp **** and then just blame the obviously incompetent machinist for being bad". Is this just a cultural difference in production industries where these is an inherent knowledge at all levels of what base requirements are needed to produce a given item? Like the board of directors knows how to run the CNC and knows the collet can't be fucked up? I imagine some dude trying to cut to 0.010 with chipped cutting tools, loose articulating arms on the CNC, keyboards that don't work, etc.
 

AEAdam

Well-known member
Joined
May 27, 2023
Messages
2,737
Location
SE PA
Is the process in your example, changing required tolerance from 30 down to 10 thou not causing higher rejection, only true if the equipment remains in a given condition and has a given output quality?
Yes sir. Spot on.
But, would this not fall off after bushings/ways wear, articulating points get sloppy, etc? Does this industry simply allow this not to happen?
Correct. This is their livelihood. So they maintain their equipment. As machines wear, quality falls off and revenue suffers. So there's a point at which, the cost to maintain or replace equipment is justified.

I think the part of this conversation that's going unsaid is, in the US when your quality fades, your name is tarnished, and that could begin your company's death spiral (SK?). In China, or Taiwan, they produce tools to get a contract with a western distributor. First articles are tested and inspected. Either intentionally or unintentionally, quality starts to slip, returns increase eventually. Could be wear, could be worker fatigue, could be intentional process/material substitution. The distributor, annoyed with their supplier, simply finds a new supplier and creates a new brand. The original supplier, anonymous to the end user, can re-tool and try again, no harm no foul. And the cycle continues. There's always another greedy distributor looking to cash in on asian made goods.
 

mike93lx

ALLIANCE MEMBER
Joined
Dec 9, 2013
Messages
37,456
Location
Richmond, VA
This is their livelihood. So they maintain their equipment. As machines wear, quality falls off and revenue suffers
Seems simple, but the reality of manufacturing has many/most companies just teetering on the edge of major mechanical failures, especially in thin margin businesses.

Snap-on's margins of about 25% at operating income are not thin at all, so they probably do take care of their stuff to a high level
 
To avoid these ads, REGISTER NOW!
Top Bottom