What's new
What's new

When machinist is just ok and they are fine with ok...

My post was not addressing the mechanical need for deviation from print tolerance but rather the administrative reason for doing so. Your close tolerance work drives a decision to go tighter...probably overseen by an owner/super/etc...to give him PEACE OF MIND that half the parts don't end up scrap...and thus the circle of life is complete...lol.;)
 
.
.
i often see a bored hole with +.0005/-.0000 tolerance but we try for +.0005/+.0003 tolerance
.
why, not all measuring tools are perfectly in calibration and calibration can easily change from heat of hand and many other factors. also hole might not be round or straight/taper bore.
.
so goal is one part fits or goes in bore. the assembler should not need to sand parts or use a hammer. part does not need to be perfect but being on extreme edge of tolerance can cause problems. that i can under stand
.
but have often seen bore measure -.0001 and trying to cut .0002 often cannot reliably be done as cutter does not want to bite in but just skate over. keep adjust .0002 and eventually cuts like .0010" or too much. sometimes part is +.0001 and in tolerance but best thing is to leave it alone as risky to go further



Now your pushing the envelope of trying to get guys to think and interpret drawings. That's done by customer or in the office.

I'm pushing for in tolerance, keeping close to the mean not hanging at the fringes of good.

What is close to the mean...take an average of highs and lows and comp by that amount. No more, no less..a good part is a good part.
 
1. Yes, I have a software development background....

2. But the issue arises independently from my personal quirks - as in "just give me the bottle of soda, I don't want a glass, I don't want ice, just hand me the bottle" and other such things in life where people are obsesses with 'doing stuff' to something that if nothing else wastes time.
 
1. Yes, I have a software development background....

2. But the issue arises independently from my personal quirks - as in "just give me the bottle of soda, I don't want a glass, I don't want ice, just hand me the bottle" and other such things in life where people are obsesses with 'doing stuff' to something that if nothing else wastes time.



LOL...customer dropped off a 3/8" plate that he wants two 1/2-13 holes. He marked where he wants them with an X using a paint marker. Marks not in center, not symmetrical...just two X's on a saw cut plate.

Told my guy to use the open bridgeport and give me two 1/2-13 on the marks.



I go past CNC mill and he's pulled vise off, has plate clamped on parallels up against pins, he's got coordinates laid out on paper to the thousandth and he's now edge finding the corner...

I asked him what the hell is he doing??

--told me it would come out more accurate this way????


yell, scream, fire or calmly do what I did.
Pull plate from machine, grab drill, put drill in chuck at Bridgeport, locate drill over the t-slot put X under drill and start drilling ...Yup "perfect location" you finish.

Less is more.
 
SIM . . . your example is perfect.

You are intent on getting the job done with the appropriate level of effort using the most appropriate means available.

Granted, we are not a job shop although we often feed a lot of business to local job shops. We do have a captive machine shop and we keep our machines busy doing one off parts for the manufacture of complex one off machines which if found to be commercially viable, create more job shop opportunity as we busy ourselves on the next prototype. . . in short this wouldn't apply to job shops that are optimized to have high volume, low mix demands with perfect drawings and no idea how the parts will be used.

As an example, we often have to mate a servo motor to a gearbox. Adapters and couplings and perhaps a prox mount to go with . . . we make 100's a year, all different. For perspective, these motor gearbox combos can often be small enough to carry in two hands, but can range in size to a few hundred HP and several thousand pounds.

Drawings get made and sent to the shop in addition to material, motors, and gearboxes being purchased. Our best machinists (as defined by making the best parts in the most efficient way and who mark up drawings and coach young engineers that we are training) . . . will often look at the job number, walk over to the job rack, and grab said servo motor and gearbox and measure those parts to see where they are with respect to tolerances and then proceed to make the parts.

Sometimes they catch a mistake on the drawings, sometimes they catch a mistake that was made when the motor or gearbox was specified incorrectly, and sometimes they catch a mistake that the motor or gearbox manufacturer has made. The key here is that a mistake is caught before it is compounded into greater expense and delay. 95 percent of the time, no mistakes, and it only takes a little more time to do this check. But for that 5 percent where mistakes are caught . . . it more than pays for the effort of all the 95 percent where no errors are found. In the end, we don't hang up a job or progress in the shop because the big picture of it having to work in an imperfect world is kept in mind.

And sometimes our machinists are tasked to make a simple set of spacers that are used to move a J-box out from a mount so that the door is flush with adjacent guarding and they know that grabbing just about any available round stock from the drop rack and parting off in the lathe to the nearest "few thou" and hitting it with a file is good enough.

The point being, they think about how the part is going to be used, and that at the end of the day it has to work for it's intended purpose, and that there is a long chain of error prone tasks that preceded them getting the print, materials, and mating parts that they might need to think about before they even think about making the part.

I realize that there are some who don't think it is the machinists job to do this quality checking . . .in our situation they are the most qualified and the most efficient at it and can often suggest the best ways to get the job done when an error is found so that we can keep our greater commitment to getting the machine built and into production.
 
@ Motion Guru. I agree with you about the machinist being "good" by checking what he is making first, EXPECIALLY in your situation. I was in a situation similar, and no one listened. I questioned why all the threads on the part were 2-56, except two at 2-64. Make it to print! So I did, then I went on the floor where parts were used, sure enough, they were jamming 2-56 screws in them. :eek: Parts were aluminum, and screws engaged maybe 3/16" into material.
I told my boss what I found and he was dumbfounded as apparently they had been doing it for years before I started, (parts used to be outsourced), I explained because the screw was so small, and aluminum material combined with a rather large screwdriver it was no problem to force the screw in.
They didn't change the threads on the part, they went with the 2-64 for those holes. Lucky people on the floor now have two part numbers to track/order and not get mixed up.:rolleyes5:
 
"You are intent on getting the job done with the appropriate level of effort using the most appropriate means available."

This is how a shop owner must think to survive. It seems much harder to transfer this....call it survival, efficiency, profit minded-ness...to a worker bee. Once again the ones that "get it" seem to "have it" pretty early on.

I am betting most of you owners and supervisors were hard wired to think this way even as lower level employees...which most of us were at some point.

I realize my stance seems somewhat negative. I think about the years I spent trying to teach people to care. Fallaciously believing that learning cool technical stuff would somehow impart integrity .Efficiency is oh-so-dependent upon integrity methinks. To truly be efficient you have to be thoughtful and that requires integrity. No-one can read minds so is the guy you tasked with a job really pouring on the coal...or thinking about beer-thirty?

If I were to do things over I would hunt for people with work ethics/drive to excel/etc and teach them the trade...rather than finding people with talent and trying to teach them to care.

most of my time spent training other people was as foreman in other peoples shops. I now use some contract labor and have a few people that are always seeking to improve the work they do for me. I am so fortunate to have found people like this! I live in an area where hard work is still a badge of honor....
 
mike1974...yep

Your story reminds me of a job (I still do occasionally)that I had done for a few years. Then I got to quote on the mold that produced the mating part. The molds are aluminum layup molds for a Kevlar part and the molds wear out regularly. But first time I got to have a go even though I was making the mating machined part for years.

As I am finishing up ten sets of molds something is niggling at me....the mold has a detail which creates a small rib on the Kevlar part....there are six of these. The machined part has eight mates to the layup part....oops.

I call the customer...they freak out...but since both prints are approved...and work fine in the real world...and been made for YEARS this way...I am told to keep making the parts. Which I have done for 6 or so years now...fortunately no human safety involved...But I still don't know if the question got raised to a review board or if I was quietly told to keep on a-goin'....

I think I know which was the path of least resistance...
 
The only thing I disagree with is "...almost out of tolerance...." . Parts are either in tolerance or out of tolerance. I have never seen (maybe they exist) an inspection report that says parts are scrap because they are almost out of tolerance. Often times appropriate use of tolerance is what makes a job profitable or not.

Just like working on framing a shed I built a few years ago. 2 other machinist type buddies helping me.....You would have thought digging a F...ing hole for concrete was brain surgery....pick up shovel, dig, repeat lol no one gives a damn if the concrete is 3" in one corner and 3.196" in the other corner.....my lawnmower weighs about 30 pounds.
 
Topics like this are one reason why SPC works to control quality. ....

Perhaps all that spc, cpk stuff goes out the window in a job shop small run.
Then is doing ok, just inside as machinist that sees the print tolerance and took the short way home.
Maybe not great but inside the print parts being shipped?
There is a print, I am inside it, not world class level but inside. What did I do wrong?
What is all this Sp, Cp, strange mathmotion stuff and why would a job shop track it or care?
Bob
 
What did I do wrong?

Nothing.

You raise a good point. None of us are doing the same work...or are the same size....or doing the same quantities, etc.

So often we are comparing apples to oranges without realizing it.....

finding common ground for discussion takes a bit of back-n-forth on these forums...and can cause disagreement when we misunderstand each others businesses.
 
If your in tolerance the job is correct thats what a tolerance is its TOLERATED.I can remember many moons ago not long finished my apprenticeship I was working on a m/c new to me and had set the job which was basically a roughing operation you roughed the job on a profile machine leaving it plus and it was finished on the next machine to size.The boss passed by and looked into the box with my completed operation on he looked at job and said did you do these I thought great the job had a fantastic finish and was mid tolerance.Yes I said proudly and he picked one up and mice,d it and said come with me and led me to the machine that was doing the finished pass on the profile.Whats the difference between your part and this one.I said that one is finished he said exactly,your part is roughed yet the finish is identical go back and increase your feed and do what the operation tells you to do.Rough the job stop wasting time trying to get mid tolerance and get good finishes.Moral of the story get in tolerance and put the required finish on.Never forgot that bollocking
 
Perhaps all that spc, cpk stuff goes out the window in a job shop small run.
Then is doing ok, just inside as machinist that sees the print tolerance and took the short way home.
Maybe not great but inside the print parts being shipped?
There is a print, I am inside it, not world class level but inside. What did I do wrong?
What is all this Sp, Cp, strange mathmotion stuff and why would a job shop track it or care?
Bob

Bob, - Agreed to a point. There are still applications in a job shop. How does a job shop owner / machinist know what his machine capability is? Wouldn't it be nice to know in a job shop that positioning tolerance for this machine is 0.XXX and for that one is 0.X1X2X3? Might make a difference in which machine he puts a part on.... or might allow him to use more tolerance and make them faster if he knows that running x close to the tolerance is OK because his machine is actually capable of still not going outside tolerance and making scrap.

Doesn't take very long to do a capability study....even in a job shop.
 
Doesn't take very long to do a capability study....even in a job shop.

Tony,
the post was sort of a stickpoke, you and I deal with stuff on larger runs and quickly learn the good and bad of it.
To a small job shop it may seem like a bunch of silly names and BS. What the heck is a Cpk, Pp or all the other weird names, and is PPAP something my wife gets?
Been a big flag waving fan of SPC from the days before there was a IBM PC.
The first computerized shop floor system I built used a Ohio Scientific with 8 dumb terminals. I did not like paper systems so well, too much effort and operator time lost along with somewhat hard to understand.
Even in a repair shop I would want to know my process capability numbers rather than some gut feel but that could be because I'm used to working with it.

On one, two, five piece runs you can and should do SPC but it's not easily supported. People miss that it's the machine and process you are charting, not the current job being run.
We all develop a gut feeling of what a machine can do.
Nice to have numbers to back it up and toss that flag that says "This machine is sick and needs attention".
Also throws the flag that says "This operator is missing something or not doing what is expected".

There is no "part is OK" crap.
Done on this machine with our methods these parts run within this. No exceptions, no surprises, no pressing the print limits.
Here is were you nail the "Part will pass inspection, it's fine" guy to the wall.
The machine guy is part of the process and much more so in manual work. Does his job or not.
Bob
 
SIM . . . your example is perfect.

You are intent on getting the job done with the appropriate level of effort using the most appropriate means available.

Granted, we are not a job shop although we often feed a lot of business to local job shops. We do have a captive machine shop and we keep our machines busy doing one off parts for the manufacture of complex one off machines which if found to be commercially viable, create more job shop opportunity as we busy ourselves on the next prototype. . . in short this wouldn't apply to job shops that are optimized to have high volume, low mix demands with perfect drawings and no idea how the parts will be used.

As an example, we often have to mate a servo motor to a gearbox. Adapters and couplings and perhaps a prox mount to go with . . . we make 100's a year, all different. For perspective, these motor gearbox combos can often be small enough to carry in two hands, but can range in size to a few hundred HP and several thousand pounds.

Drawings get made and sent to the shop in addition to material, motors, and gearboxes being purchased. Our best machinists (as defined by making the best parts in the most efficient way and who mark up drawings and coach young engineers that we are training) . . . will often look at the job number, walk over to the job rack, and grab said servo motor and gearbox and measure those parts to see where they are with respect to tolerances and then proceed to make the parts.

Sometimes they catch a mistake on the drawings, sometimes they catch a mistake that was made when the motor or gearbox was specified incorrectly, and sometimes they catch a mistake that the motor or gearbox manufacturer has made. The key here is that a mistake is caught before it is compounded into greater expense and delay. 95 percent of the time, no mistakes, and it only takes a little more time to do this check. But for that 5 percent where mistakes are caught . . . it more than pays for the effort of all the 95 percent where no errors are found. In the end, we don't hang up a job or progress in the shop because the big picture of it having to work in an imperfect world is kept in mind.

And sometimes our machinists are tasked to make a simple set of spacers that are used to move a J-box out from a mount so that the door is flush with adjacent guarding and they know that grabbing just about any available round stock from the drop rack and parting off in the lathe to the nearest "few thou" and hitting it with a file is good enough.

The point being, they think about how the part is going to be used, and that at the end of the day it has to work for it's intended purpose, and that there is a long chain of error prone tasks that preceded them getting the print, materials, and mating parts that they might need to think about before they even think about making the part.

I realize that there are some who don't think it is the machinists job to do this quality checking . . .in our situation they are the most qualified and the most efficient at it and can often suggest the best ways to get the job done when an error is found so that we can keep our greater commitment to getting the machine built and into production.

Motion, you have very good machinists. I have Tool & Die Makers with quite a bit of skill. Taking the time to explain the purpose of what they are making eliminates a ton of mistakes. When an experienced guy looks at something, and thinks, "That doesn't look right." or, "Something is wrong here." and they don't know why; the experienced hand will shut the machine off and find out why he feels that way. At least the good ones will. Hell, the good ones deservedly ride me hard when I make a mistake. I guess I'm saying that the more information in the Maker's hands, the better the part will be.

That said, in tolerance is in tolerance. Also, the old saw about we specialize in time, money, and accuracy, you can choose which two of those you'd like, applies.
 
The only thing I disagree with is "...almost out of tolerance...." . Parts are either in tolerance or out of tolerance. I have never seen (maybe they exist) an inspection report that says parts are scrap because they are almost out of tolerance. Often times appropriate use of tolerance is what makes a job profitable or not.

Just like working on framing a shed I built a few years ago. 2 other machinist type buddies helping me.....You would have thought digging a F...ing hole for concrete was brain surgery....pick up shovel, dig, repeat lol no one gives a damn if the concrete is 3" in one corner and 3.196" in the other corner.....my lawnmower weighs about 30 pounds.

I hear you. Inspector here likes that line. :rolleyes5:
him - this dimension runs high sometimes
me - how much?
him - .002"
me - we have .005". it is still well within (silently cussing him)
him - well, can't we try to move it a little?
me - no, it is such and such tool, blah blah

It's not that I am lazy, it was just a pain in the ass to fix kind of thing. I could understand if it was .002" over, then every 5-10 parts it would be .0035-.004" over. OK, that is a problem that needs addressed. A history of parts that float between nominal and plus .002" does not need addressed in my opinion.
 
Tony,

............

On one, two, five piece runs you can and should do SPC but it's not easily supported. People miss that it's the machine and process you are charting, not the current job being run.
We all develop a gut feeling of what a machine can do.
Nice to have numbers to back it up and toss that flag that says "This machine is sick and needs attention".
Also throws the flag that says "This operator is missing something or not doing what is expected".

There is no "part is OK" crap.
Done on this machine with our methods these parts run within this. No exceptions, no surprises, no pressing the print limits.
Here is were you nail the "Part will pass inspection, it's fine" guy to the wall.
The machine guy is part of the process and much more so in manual work. Does his job or not.
Bob

Highlighted that line above...... people really, really need to understand that.

What most don't realize is that it also reduces your need for inspection and thus saves you TIME. TIME is your most important metric in a job shop, that is what you are actually selling. If your machine and process is proven to repeat within 10% of the tolerance and you center it in the tolerance band, you shouldn't be doing 100% inspection on that feature. First, middle and last of the run is sufficient. The likelihood of it running a bad part is very small. Guy running the machine can be doing something more productive than measuring that feature.
 
Highlighted that line above...... people really, really need to understand that.

What most don't realize is that it also reduces your need for inspection and thus saves you TIME. TIME is your most important metric in a job shop, that is what you are actually selling. If your machine and process is proven to repeat within 10% of the tolerance and you center it in the tolerance band, you shouldn't be doing 100% inspection on that feature. First, middle and last of the run is sufficient. The likelihood of it running a bad part is very small. Guy running the machine can be doing something more productive than measuring that feature.

I know you deal with this stuff on a daily basis, so I will take your word for it.

Don't you have to do inspections per your sampling plan (not sure the term)? So for a *1.0 D, which is 26-50 pieces, the sample size is 8.

*ref ANSI/ASQC Z1.4-1993
 
I know you deal with this stuff on a daily basis, so I will take your word for it.

Don't you have to do inspections per your sampling plan (not sure the term)? So for a *1.0 D, which is 26-50 pieces, the sample size is 8.

*ref ANSI/ASQC Z1.4-1993

In our case, we sample based on the historic/ongoing Cpk/Ppk and yearly Cm/Pp testing. We do sample much more frequently (similar to ANSI but the ISO equiv) for smaller lots (<600 pcs), but again, only for features that have <3.0 Cpk, otherwise, it is first, middle, last and tool changes. Having solid documentation of your process capability can reduce your need for inspection. We have some processes for features where the Ppk is greater than 8. That means something on the order of 1 bad part possible in 20 or 30 million (someone with minitab might check that for me) Why would you inspect that feature other than first (verify set-up), tool change (again, verify set-up), middle and last piece. Any further inspection is a waste of time and resources.
 








 
Back
Top