But the guy said “meet the exact specifications on the blueprints.” That doesn’t mean the size must be “perfect,” it means it must meet tolerances specified in the blueprints. It’s not the machine shop’s job to judge which tolerances can be ignored.
Exquisite timing 
Sounds like management doesnt understand what a “perfect part” is. I’ve done this from both the machinist side and the engineering side. If you have a dimension that is +/- .002, that’s a critical dimension. I even run a CNC grinder sometimes holding tolerances of +/-.0003. That tolerance is for a press fit, for a bearing that will be pressed on.
Each tolerance has a meaning. .0003 (called 3 tenths) is a press fit like I said. .002 (simply called “2”) is for a slip fit, like a shaft that has to move back and forth thru a hole - loose but a precision location. +/- 10 or even 20 is just a clearance dimension. Like if you were making a hex bolt. The size of the hex would be held fairly close (+/- 5 or so) so that the wrench fits it. But the length of the hex would be +/- 20. It doesn’t matter at all (except in very specific apps, and then they would put a different tolerance on it…) A +/- 20 dimension really only has a tolerance associated with it because engineers and draftsman have to put a tolerance on it. Or it is generically called out on the title block of the drawing.
In other words, a “perfect part” is one on which the critical dimensions are held, and works in the application for which it was designed. Your manager doesn’t seem to realize this. If he’s scrapping parts over non-critical dimensions, or wasting hours of his machinist’s time trying to fix this, or just make “pretty” parts, he’s wasting a LOT of his company’s money and resources. Management for the sake of management, not for any tangible reason.
True, but as we switch over to more and more aviation components, the tolerances we have to hold are tighter than what the machines are capable of holding. Even worse, because the machines are improperly maintained, getting them to hold any tolerance is difficult simply due to thermal expansion. They won’t clean the chips out of the machines, these chips displace the cutting oil we use to keep the tools cool while they cut, the chips also hold heat. Some of the parts we run are so hot when they come out of the machines that you can’t grab them with your bare hands. You have to grab the part with a shop towel and blast it with an airline before you can check it.
if6was9, you’re exactly right. It doesn’t matter to anyone but him if the countersink on one side is .050 while the one on the opposite side is .060 and the print calls for them to be .030 - .050. It’s not something that anyone’s going to notice unless they look closely at it, it’s not going to affect how the nut holds the bolt, and so it’s not something you should sweat that much. What I worry about is things like one of my cow-orkers does and uses whatever size drill he thinks should drill the hole, rather than a drill which makes a hole the size which fits the gauge. His parts get scrapped a lot because of this, but no one disciplines him because of it.
Are you in a position to set up a preventative maintenance program? You sound like you know what would need to be done on a lot of your machines, so drafting even a rough set of “weekly, monthly, 6 months, yearly” actions to take to ensure that the machines don’t fail is already a huge step forward. You might not be able to replace the machines with newer models, but at least you can do what you can to make what you have work longer and better.
I don’t really understand your last example, where one side is 0.050 and the other 0.060… that last measurement is out of spec, and therefore unacceptable. WHY is the spec 0.030 to 0.050? Is there an industry precedent for that? Some sort of regulation/standard for it to be based upon? Perhaps taking the time yourself to see if the specs can be modified (say to 0.040 to 0.060) might be worth it, if that’s consistently the product that you produce and it can be shown that it doesn’t affect your clients.
As for paperwork… gah, I know what it is to do a lot of meaningless paperwork, but the one time someone bothers to ask the question, and you have zero documentation on something, is the one time you get seriously screwed! Again, it could be a lot of work for you, but if you rewrote the forms with the relevant things to be filled out, and cut out the useless crap, then perhaps you could get your supervisors/managers/QA to approve it?
I’m in a bit of a similar situation at my job, where I find that they are overlooking a lot of safety issues, and in a few cases, frankly, I think they are calculating results/running tests incorrectly. I’ve tried to simply talk about it, but it’s pretty much getting to the point where I will have to write a freaking manifesto to give to my supervisor/manager/VP in order to get them to think about what I’ve said. I’m sort of addressing things one thing at a time, but it is SO frustrating. And what’s worse, IMHO, is they very recently caved in to some fucked up idea an auditor had, and changed our SOP despite the fact that everyone knows it’s wrong, rather than standing up to the auditor and going through the appeal process that would prove that the guy is a dipshit. I mean, seriously, it IS in fact possible for something to be greater than 100% when it is being compared to something else… that something else might, really, not be as pure as you assume it is! ARRGGHH!
No. Were I to attempt such a thing, I would quickly found myself fired. Management does not see a reason to do preventative maintence.
The spec is that way, because that’s how the engineer designed it. It may be an industry standard, it may be a custom request by a customer. It all depends upon the final application which determines what the specs will be.
Machine A might produce a part at those specs, but machine B won’t, it will produce them at some other spec. IAC, it doesn’t really matter because so long as the countersink is above a certain size, or doesn’t wipe out the bearing surface, it’s actual size will have no affect on the performance of a nut.
Trust me, no one uses the “information” on the paperwork. It proves nothing.
Bwahhhaaaa! At best, I’d get a pat on the head and told what a good boy I was, but I should leave the important thinking to people who know what they’re doing, at worst, I’d suddenly find every minor mistake I made documented and then promptly be fired for “poor job performance.”
This is why God invented OSHA.
And what makes you think they’ll bother to do anything, muchless read it? If they’re not willing to listen to you when you talk to them, how is paperwork going to change their mind? Unless, of course, you plan on submitting copies of that paperwork to the appropriate government body?
Been there, done that. What you need to solve the problem is a Very. Big. Guy. Have Very. Big. Guy. explain the problem to the boss rather energetically. Very. Big. Guy. should be animated, but polite, while looming largely over the boss. Boss will soon realize that if he doesn’t make some changes, Very. Big. Guy. might just rip his head off. Boss will have changes made immediately. I’d do it where I work, but there’s no one bright enough or large enough to be able to do it.
Tuckerfan, sounds like you’re cursed with being in a dying company. I’d bet the stockholders don’t really even expect it to last too long, but are just trying to squeeze the last few drops of blood from the turnip. I’ve worked in companies like that, too, (but in a different industry); once they get into that cycle of throwing new managers without relevant experience at the problem, but expecting instant results, it’s hard to turn things around. Atari was like that in the early 80’s. After Bushnell left, they brought in a guy from Burlington Industries (the textile manufacturer) – he got along OK because video games were just going gangbusters and he could just throw money around without anyone noticing how much of it was being wasted. Then things got tough – so they brought in a guy whose entire career was in marketing tobacco products – no knowledge of electronics, entertainment, or Silicon Vally culture. That went well.
The tolerances stuff reminds me of the summer I worked for a machine shop at a major research university. The shop made custom experimental apparatus, and most of the machinists had decades of experience. They were always laughing at the physics grad students, some of whom had hardly used so much as a screw driver, coming in and insisting that their parts had to be “exact” and made with “zero tolerances.” Though they’d become accustomed to it over the years, it was a constant frustration educating their clients on what was possible with the available equipment and the budget available.
I think Tuckerfan probably knows what he’s talking about. As nice as it is to hand out advice, I thin I understand where he’s coming from. His bosses KNOW they’re doing things right, and no amount of reason will ever convince them otherwise. It’s pretty much pointless to actually try and solve the problems, because they refuse to acknowledge that any problems exist except in the employees.
Heh. Did the physics students specify frictionless materials and perfect insulators, too? 
I can just imagine a few of them trying out their latest experiments. “It’ll prove I’m an absolute genius, just as long as that damned machine shop can violate all materials science.”
Sadly, the physicists I’ve met are so woefully imcompetant (and often unwilling to learn) that I just roll my eyes at them.
No, actually these students were quite brilliant, within their field. But all those QM didn’t leave them a lot of time to study mechanical design.
Funny, when I was a mechanical engineering undergrad, I was one of the few guys in my department who didn’t spec everything as “.0001” on all tolerances. I had one part that had two very precise holes (where two square shafts needed to be inserted). Those two were marked “press fit for soft aluminum” and all other tolerances were 0.1" – funny how my parts always seemed to be done so fast!
Tuckerfan, you don’t happen to work for a publicly-traded company, do you?
OH! I know all about this line. Bog used to work the job you’re doing. I heard earfuls on crap like this. Decrees from on high about how those things on the floor will go. My favorite was the time estimates given to customers. Like: Sure! We can give you 100,000 parts in 2 days. Then they’d go to Bog and say, “Produce 100,000 parts in 2 days”. “Um… First, the machine won’t run that many in that time.” “Sure it will! It can run 50,000 per day.” “Um. Yeah. If running 24x7 with EVERY single part coming out perfect, never needing to stop it to add materials, switch operators, etc.” These people had no clue what the real production amounts were. But they’d expect him to do it anyway. Like he was just trying to wriggle out of doing any work.
Oh… And the obligatory… “Fuckwits!”
Please don’t lump experimentalists with theorists. Most experimental scientists (including myself) are perfectly capable of looking up thermal and mechanical properties of available materials and doing a numerical analysis of our instruments.
Well, even the theorists know that frictionless materials and perfect insulators aren’t available.
But few undergrads get to design apparatus that require the use of a machine shop. When they get to the research stage of their grad program, many of them are learning about mechanical design for the first time.
Thing is, many times tolerances are badly defined. That part tuckerfan mentioned, where it’s OK to send it .005 undersize but not .005 oversize, should either not be defined as having a tolerance of “+/- .005” (have more tolerance on - than on +) or the center of the range should be moved. Dimensions should have their tolerances set to the values that QM and the customer really will accept. One of the things I’ve had to do as part of my work is get people to give me a “believable upper limit” because they wanted to use a single, lower limit and I knew they’d rejected stuff for having that value too high (particle size for an emulsion; too high means clogged machine parts).
Sometimes this is hard to arrange with the customers, since “our” guys aren’t the only ones who wouldn’t recognize a t function if it hit them with a hammer, but many times all it takes is one phone call and two minutes of typing. Oh, yeah. And common sense, sigh. And the daring required to call the customer and say “you know, I was wondering, would you have a problem if we change this number to this other value?”
The place where I’m currently working is going to use every single function that the program can provide. About 50% of it corresponds to things they didn’t consider necessary until they heard the new program can do it. The time sheets (one of the new functions) are going to say that everybody did everything in exactly the expected time. The expected time is going to be invented, since nobody is sure how long things take. I’m kind of looking forward to checking out, a few months from now, whether anybody has ever pulled out a time sheets report. My bet, based on previous projects, is “not”.
Admittedly, I never asked them what their speciality and actual fields of study were. I just remember that every bloody physicist I ever met in person was a total space case with no knowledge of anything outside their offices. Down to the idiot who’d been using Windows for a decade and still had to be taught how to use every basic function, down to me specifying which mouse buttons to use. He simply refused to learn anything.
And the other idiot who got a job and walked off in the middle of it, leaving someone else to pick up the shattered peices of the home surveys.
Come to think of it, though, I’d have probably done the same thing since the manager wanted to pay him 20K for random 8-12 hr days with constant business trips to miserable small towns. 
But how do you KNOW? You said the inspection equipment is just as crappy as the manufacturing equipment.
In my department, some of the inspection equipment is crappy. Calipers aren’t an accurate way of measuring things, but thread gauges are. Also, they’ve never tried to substitute the incorrect material on an order, and the final inspection equipment all has to be highly accurate, certified, the readings from those devices is recorded, signed off by the inspector, and submitted to the customer when the parts are shipped off. If a component were to fail in the field it is that paperwork and equipment which would be under scrutiny by the Feds. Finally, I have my own calipers, micrometers, etc. and I use that whenever I can to check the parts. I’ve checked the differences between what my stuff says and what the company’s stuff says, and there’s not enough of a difference in the readings to cause concern, however, there’s enough of a difference that you can get written up for producing bad parts.
That’s a relief. I was concerned you (the corporate “you”) were leaving it to your customers to catch.
Well, there’s a big meeting scheduled for tomorrow, and apparently, the division head is going to be there. This is not a good thing in every one’s opinion. El Presidente is really screwing things up, too. I had a number of employees tell me that they’re planning on looking for new jobs because of his policies, many of them having been with the company for a decade or more. He also chewed the shit out of one of the supervisors for having his personnel working in an area that the president didn’t think that they should be in. The supervisor had to point out that there was no work available in the other area, so he had to put everyone in the one area.
It’ll be interesting to see if I have a job after tomorrow.
As usual, there is a Dilbert cartoon for every work situation. I can’t remember the specifics, but basically the “evil boss” character estimates a timeline for Dilbert’s project on the assumption that any task he doesn’t understand can be done easily and in minimal time.