- Joined
- Mar 30, 2023
story time
This is something I need to get off my chest, but I haven't been able to talk about it to someone who knows machining well enough to understand how bad it's gotten at times.
I started my last job after being off work for a year during covid lockdowns, I had just finished my apprenticeship and became a very red seal machinist before I was let go from my old job. I felt lucky to get the job, as I was looking for some time and had gotten the phone call after literally given up and was about to visit family in a small town and see if I could get odd jobs and rent a place there. The shop itself was impressive, with a couple fleets of six axis lathes with live tooling and dual turrets, and a customer list that included some big names. When doing the tour I spotted an area where they had cut out the concrete pad, which turned out to be important.
A month or two into my probation I was asked to train on their brand new machine, a horizontal milling machine that could retract its spindle to use gundrills with for deep hole drilling. This was a massive machine - a Cheto CCN - where the table rotated and moved on the z axis, and the spindle moved on the x and y axis, and was enclosed in a giant box you could walk in to access the parts. Management was excited for the machine as they bought it for this complicated project from one of their bigger customers, with the expectation that the project would pay for the machine. In hindsight, that should've been a warning that the project would be complicated.
I show up for training on the machine and met the guy they flew in from Portugal. He was standing around confused, as the company had not finished (or even started) the fixture needed for the project he was training us to run, and not all of the tooling has arrived. As it turns out he was more of a programmer then a trainer because all of their trainers were stuck in China due to mandatory covid isolation during travels, so he was all we got. I was training along with two other guys - one who was pulled away one day and I never saw again, and the other was the night time foreman and was busy with the shop half the time. The Forman never completed the remaining either, so I was the sole person operating the machine.
Then I learned what the actual process was, and slowly learned how badly things were planned. In this company, the main owner was a foreman before and knew the machining process well enough to actively advice when needed, problem was that his kids were running the things. They were probably nice enough, but they went straight from management school to running the company without ever touching metal. All their old machines were from the same supplier that has worked with them before and their process with dedicated programmers and qc was pretty much figured out, but this machine was uncharted territory, and everything that could go wrong, did.
First, the machine itself. When buying the machine they had a choice of a more expensive machine from an American supplier, but they chose the much cheaper machine from a Portuguese company that has only shipped three machines total to America, and each one had to be manually reprogrammed to work in inches. Everything on that machine was metric or European, and the littlest things threw you. The included grinder needed a special plugin, and the literal air hose kept popping off until we realized we needed a metric sized hose in that fitting. And a new post had to be written from scratch for our cam software, and to my knowledge it has never been completed as the cam developers in California had very poor communication with the cnc developers in Portugal (and every conversation required 24 hours as the work hours were complete opposites). I ended up writing all the programs myself, relying heavily on the inserts' recommended speeds and feeds for inconel.
Their initial plan was to put the parts in a fixture in the middle of the table so they could machine one side, rotate the table 180 degrees, then do the other side. Perfectly logical in theory.
First: these parts were inconel, rings about six inches in diameter and four inches long, with a wall thickness of about 0.5-1". They got six 1/4" through holes with cointersinks on either end, one with a tolerance of 0.0002" for diameter, and hole alignment requirements of half a thou. Oh and a massive groove on the inside bore that intersects the holes, so it must be done last. So very fussy parts, and also parts that we eventually figured out how to make on the fleet of lathes we have (for faster, but we'll come back to that).
With the setup planned, the spindle doesn't even come close to reaching the parts for the millong operations (this machine is a gundrill first, horizontal mill secondary). Thus the plan was to have 12" extensions for all of the tool holders, with a even longer one for the 50mm face mill used to cut the groove. Management didn't see any problems with that when they bought the machine. They didn't stop to wonder if hitting a 0.002" tolerance in inconel might be tricky, nor if the spindle of a gundrill would handle an oversized overlong facemill taking heavy cuts at low rpms well.
Did I mention that management insisted on doing the first few parts using the customer's material instead of cheaper steel or aluminum?
Eventually management concluded that the extensions were a bad idea, and agreed to have the fixture moved to the edge of the table so the tools could be shortened up as much as possible, with the plan of physically flipping the fixture to do the other side. In theory this would be quick and easy, as they bought a fancy fast clamp system that the fixture sat on which aligned it perfectly each time. Except the fixture itself wasn't well thought out.
Let me describe the fixture as follows. Imagine a four foot long block of steel about four inches wide and six inches tall, with a flared base six inches wide and four v-block grooves in a row to hold each part. Bolted to the back of it is a 1" thick L-shaped panel spanning the width of the fixture, designed to keep the parts in line for the first operation but must be removed for machining the back sides of the parts. I dont recall the exact weight of this panel, but it is heavy enough that most people will have difficulty taking it on or off (I eventually chopped it in half which helped a lot). So now you have four v-grooves to hold four parts and a panel to align them with, but how do you strap them down? The solution was a 4x6x48" block with identical grooves, which requires a crane to maneuver. To hold thin walled parts with an outside diameter tolerance of 0.015".
Management saw no issues with this design.
Let's set aside the fact that this fixture did not get made until several weeks after training and testing the new machine was supposed to occur. We will start with this: the blueprints for this fixture (which admittedly was helpfully converted from metric to imperial units) had absolutely zero GD&T callouts, something everyone missed until I go to use the fixture and discovered that the flat faces of the V grooves were not straight in relation to any surface known to man. The machinist that milled the thing didn't notice, and QC signed off on it as within tolerances, but when I go to put a part in each part is crooked in a way I can't compensate for. Oh and fun fact: if a round part has thin walls, even if it is inconel, it will squish if you try to clamp it to a v-block. Not ideal with tight tolerances.
I'm running out of steam for my rant so I'll try to summarize. Eventually we bored round holes in the fixture to hold the parts, remade the parts with a much higher od tolerance, and eventually resorted to only using the leftmost bore of the fixture to hold the part. Oh and added an insert into the part that helped it keep its shape. And changed the order of operations to do the back side first, as doing that big groove in the middle always spun the part in the fixture a smidge.
The entire process was hell. I was the only person operating the machine, or even touching it. When a spindle sensor failed (which became a reoccurring theme), I had to stay after work a couple hours just twiddling my thumbs because maintenance required me to enter a specific M-code to test it, something I could've shown them or basically anyone how to do (possibly a trained monkey if you could ever trust one to operate a mill) but the machine scared them too much.
We scrapped so many parts trying to dial it in. The initial plan was they could do four parts in a day, but to hit the high tolerances we had to slow down some operations a LOT. I've come to realize that the machine simply wasn't ridgid enough for our needs, which explains the floor needed to be leveled as its too flimsy for levelling jacks. We started with customer material (which they were happy with) before switching to cheaper steel until we finally got the process dialled in, then scrapped the first few inconel pieces again for good measure.
And every single day management dropped by telling me that they need the part down that night as they promised the customer it would be shipped to them tomorrow (to the extent that I had a couple illegal 14- hour shifts to get the job done), only to come back to work the next day to find out the cmm determined it was scrap for a slightly different reason. The rational thing was to recognize we were having problems and negotiate a delay with the customer so we could take the time to reassess and do it right, but instead we spent a couple months promising the customer we would get them one part the next day, only for the part to be scrapped because we rushed or got hacky with temporary fixes so we had to delay it yet another day.
We eventually got it working, but not before getting the customer to approve a poorer quality finish on the through hole (the whole point of using the gundrill) so we could use our existing fleet of lathe's with life tooling and dual spindles. We do make some parts on the gundrill though, when the regular fleet is busy and we need to justify the expensive paperweight of a machine. We did lose the job when the customer got pissed off enough with the delays (which, again were "it'll be done tomorrow pinky swear" repeated indefinitely instead of "we are having issues with the process and so need to delay for a couple weeks to order new tooling and figure things out", that latter of which in my untrained opinion seems more professional and a customer would be able to schedule around.
After that I floated around the fleet when we didn't have that job for a while. Six months later we got a second job for the gundrill which slowly turned out to be another nightmare (we hadn't clued into the machine's lack of ridgidity, and couldn't get the gundrill holes straight enough) and I had a full on mental breakdown to the point I was physically unable to go to work for weeks (solely surviving because I caught covid and at the time the government paid me to stay home if I had a positive test). I did go back when the covid cash ran out, but my attendance was spotty. I tried for the last couple months trying to find work outside of machining before caving in and applying to a much nicer shop. I was still broken and eventually was let go, which let me go on EI and do a retraining program to work at an office. I still have some hope in this new career path.
The company was fucked up. Because of the shiny epoxy coated floors oil from the machine collected on them and became a major skipping hazard. I had requested a mop and bucket of my own as there's only one at the shop and it's used for the offices, and it took me almost a year to get it because they refused to spend money on important little things like that. After a couple months of me complaining they wanted to do something fancy and got disposable carpet, which is a long roll of felt that has adhesive to keep it in place. It worked for a few weeks, but then science happened; the oil seperated the felt from the adhesive and bonded to the adhesive and fused it to the floor. That made a surface that was very very slippery and dangerous, and it could not be washed off even with shop-grade cleaners. I tried everything we had, and even resorted to borrowing a bit of goo gone from home. But nooo they refused to buy enough goo gone to get teh stuff off, and even then it didn't work the greatest. Eventually I discovered I could scour the gunk off with the tri-dry oil absorbent and got it off with serious elbow grease. After that I finally got the mop and bucket I requested a year ago Q's they bought three sets for the shop, but they got the cheapest ones they could and they broke within a month.
Plus they broke my toolbox. I came to work one day and found the drawers all fucked up and the contents a mess, and when I asked around it turns out that it got hit by a forklift and knocked over (this was a cheap toolbox that I kept locked as it had no latches but still) and the rails for the bottom drawer were ruined. I pointed it out to my job leader and asked if they would replace it and guess what he said? Yes they would. Except they didn't. I spent months asking him about it and emailing him about it and even found an equivalent one that was like $300 that they agreed to buy to replace it, but they never ordered it and he would make a big show of having forgotten it and promised to order it that night every single time. Eventually I stopped asking but by then I had nmny breakdown and couldn't afford to replace it. I would have been disappointed but would've accepted if they flat out told me they weren't responsible for the damage, but to yank my chain around for nine months with promises they very carefully avoided ever putting in writing.
Its been years and I'm still angry at that company.
This is something I need to get off my chest, but I haven't been able to talk about it to someone who knows machining well enough to understand how bad it's gotten at times.
I started my last job after being off work for a year during covid lockdowns, I had just finished my apprenticeship and became a very red seal machinist before I was let go from my old job. I felt lucky to get the job, as I was looking for some time and had gotten the phone call after literally given up and was about to visit family in a small town and see if I could get odd jobs and rent a place there. The shop itself was impressive, with a couple fleets of six axis lathes with live tooling and dual turrets, and a customer list that included some big names. When doing the tour I spotted an area where they had cut out the concrete pad, which turned out to be important.
A month or two into my probation I was asked to train on their brand new machine, a horizontal milling machine that could retract its spindle to use gundrills with for deep hole drilling. This was a massive machine - a Cheto CCN - where the table rotated and moved on the z axis, and the spindle moved on the x and y axis, and was enclosed in a giant box you could walk in to access the parts. Management was excited for the machine as they bought it for this complicated project from one of their bigger customers, with the expectation that the project would pay for the machine. In hindsight, that should've been a warning that the project would be complicated.
I show up for training on the machine and met the guy they flew in from Portugal. He was standing around confused, as the company had not finished (or even started) the fixture needed for the project he was training us to run, and not all of the tooling has arrived. As it turns out he was more of a programmer then a trainer because all of their trainers were stuck in China due to mandatory covid isolation during travels, so he was all we got. I was training along with two other guys - one who was pulled away one day and I never saw again, and the other was the night time foreman and was busy with the shop half the time. The Forman never completed the remaining either, so I was the sole person operating the machine.
Then I learned what the actual process was, and slowly learned how badly things were planned. In this company, the main owner was a foreman before and knew the machining process well enough to actively advice when needed, problem was that his kids were running the things. They were probably nice enough, but they went straight from management school to running the company without ever touching metal. All their old machines were from the same supplier that has worked with them before and their process with dedicated programmers and qc was pretty much figured out, but this machine was uncharted territory, and everything that could go wrong, did.
First, the machine itself. When buying the machine they had a choice of a more expensive machine from an American supplier, but they chose the much cheaper machine from a Portuguese company that has only shipped three machines total to America, and each one had to be manually reprogrammed to work in inches. Everything on that machine was metric or European, and the littlest things threw you. The included grinder needed a special plugin, and the literal air hose kept popping off until we realized we needed a metric sized hose in that fitting. And a new post had to be written from scratch for our cam software, and to my knowledge it has never been completed as the cam developers in California had very poor communication with the cnc developers in Portugal (and every conversation required 24 hours as the work hours were complete opposites). I ended up writing all the programs myself, relying heavily on the inserts' recommended speeds and feeds for inconel.
Their initial plan was to put the parts in a fixture in the middle of the table so they could machine one side, rotate the table 180 degrees, then do the other side. Perfectly logical in theory.
First: these parts were inconel, rings about six inches in diameter and four inches long, with a wall thickness of about 0.5-1". They got six 1/4" through holes with cointersinks on either end, one with a tolerance of 0.0002" for diameter, and hole alignment requirements of half a thou. Oh and a massive groove on the inside bore that intersects the holes, so it must be done last. So very fussy parts, and also parts that we eventually figured out how to make on the fleet of lathes we have (for faster, but we'll come back to that).
With the setup planned, the spindle doesn't even come close to reaching the parts for the millong operations (this machine is a gundrill first, horizontal mill secondary). Thus the plan was to have 12" extensions for all of the tool holders, with a even longer one for the 50mm face mill used to cut the groove. Management didn't see any problems with that when they bought the machine. They didn't stop to wonder if hitting a 0.002" tolerance in inconel might be tricky, nor if the spindle of a gundrill would handle an oversized overlong facemill taking heavy cuts at low rpms well.
Did I mention that management insisted on doing the first few parts using the customer's material instead of cheaper steel or aluminum?
Eventually management concluded that the extensions were a bad idea, and agreed to have the fixture moved to the edge of the table so the tools could be shortened up as much as possible, with the plan of physically flipping the fixture to do the other side. In theory this would be quick and easy, as they bought a fancy fast clamp system that the fixture sat on which aligned it perfectly each time. Except the fixture itself wasn't well thought out.
Let me describe the fixture as follows. Imagine a four foot long block of steel about four inches wide and six inches tall, with a flared base six inches wide and four v-block grooves in a row to hold each part. Bolted to the back of it is a 1" thick L-shaped panel spanning the width of the fixture, designed to keep the parts in line for the first operation but must be removed for machining the back sides of the parts. I dont recall the exact weight of this panel, but it is heavy enough that most people will have difficulty taking it on or off (I eventually chopped it in half which helped a lot). So now you have four v-grooves to hold four parts and a panel to align them with, but how do you strap them down? The solution was a 4x6x48" block with identical grooves, which requires a crane to maneuver. To hold thin walled parts with an outside diameter tolerance of 0.015".
Management saw no issues with this design.
Let's set aside the fact that this fixture did not get made until several weeks after training and testing the new machine was supposed to occur. We will start with this: the blueprints for this fixture (which admittedly was helpfully converted from metric to imperial units) had absolutely zero GD&T callouts, something everyone missed until I go to use the fixture and discovered that the flat faces of the V grooves were not straight in relation to any surface known to man. The machinist that milled the thing didn't notice, and QC signed off on it as within tolerances, but when I go to put a part in each part is crooked in a way I can't compensate for. Oh and fun fact: if a round part has thin walls, even if it is inconel, it will squish if you try to clamp it to a v-block. Not ideal with tight tolerances.
I'm running out of steam for my rant so I'll try to summarize. Eventually we bored round holes in the fixture to hold the parts, remade the parts with a much higher od tolerance, and eventually resorted to only using the leftmost bore of the fixture to hold the part. Oh and added an insert into the part that helped it keep its shape. And changed the order of operations to do the back side first, as doing that big groove in the middle always spun the part in the fixture a smidge.
The entire process was hell. I was the only person operating the machine, or even touching it. When a spindle sensor failed (which became a reoccurring theme), I had to stay after work a couple hours just twiddling my thumbs because maintenance required me to enter a specific M-code to test it, something I could've shown them or basically anyone how to do (possibly a trained monkey if you could ever trust one to operate a mill) but the machine scared them too much.
We scrapped so many parts trying to dial it in. The initial plan was they could do four parts in a day, but to hit the high tolerances we had to slow down some operations a LOT. I've come to realize that the machine simply wasn't ridgid enough for our needs, which explains the floor needed to be leveled as its too flimsy for levelling jacks. We started with customer material (which they were happy with) before switching to cheaper steel until we finally got the process dialled in, then scrapped the first few inconel pieces again for good measure.
And every single day management dropped by telling me that they need the part down that night as they promised the customer it would be shipped to them tomorrow (to the extent that I had a couple illegal 14- hour shifts to get the job done), only to come back to work the next day to find out the cmm determined it was scrap for a slightly different reason. The rational thing was to recognize we were having problems and negotiate a delay with the customer so we could take the time to reassess and do it right, but instead we spent a couple months promising the customer we would get them one part the next day, only for the part to be scrapped because we rushed or got hacky with temporary fixes so we had to delay it yet another day.
We eventually got it working, but not before getting the customer to approve a poorer quality finish on the through hole (the whole point of using the gundrill) so we could use our existing fleet of lathe's with life tooling and dual spindles. We do make some parts on the gundrill though, when the regular fleet is busy and we need to justify the expensive paperweight of a machine. We did lose the job when the customer got pissed off enough with the delays (which, again were "it'll be done tomorrow pinky swear" repeated indefinitely instead of "we are having issues with the process and so need to delay for a couple weeks to order new tooling and figure things out", that latter of which in my untrained opinion seems more professional and a customer would be able to schedule around.
After that I floated around the fleet when we didn't have that job for a while. Six months later we got a second job for the gundrill which slowly turned out to be another nightmare (we hadn't clued into the machine's lack of ridgidity, and couldn't get the gundrill holes straight enough) and I had a full on mental breakdown to the point I was physically unable to go to work for weeks (solely surviving because I caught covid and at the time the government paid me to stay home if I had a positive test). I did go back when the covid cash ran out, but my attendance was spotty. I tried for the last couple months trying to find work outside of machining before caving in and applying to a much nicer shop. I was still broken and eventually was let go, which let me go on EI and do a retraining program to work at an office. I still have some hope in this new career path.
The company was fucked up. Because of the shiny epoxy coated floors oil from the machine collected on them and became a major skipping hazard. I had requested a mop and bucket of my own as there's only one at the shop and it's used for the offices, and it took me almost a year to get it because they refused to spend money on important little things like that. After a couple months of me complaining they wanted to do something fancy and got disposable carpet, which is a long roll of felt that has adhesive to keep it in place. It worked for a few weeks, but then science happened; the oil seperated the felt from the adhesive and bonded to the adhesive and fused it to the floor. That made a surface that was very very slippery and dangerous, and it could not be washed off even with shop-grade cleaners. I tried everything we had, and even resorted to borrowing a bit of goo gone from home. But nooo they refused to buy enough goo gone to get teh stuff off, and even then it didn't work the greatest. Eventually I discovered I could scour the gunk off with the tri-dry oil absorbent and got it off with serious elbow grease. After that I finally got the mop and bucket I requested a year ago Q's they bought three sets for the shop, but they got the cheapest ones they could and they broke within a month.
Plus they broke my toolbox. I came to work one day and found the drawers all fucked up and the contents a mess, and when I asked around it turns out that it got hit by a forklift and knocked over (this was a cheap toolbox that I kept locked as it had no latches but still) and the rails for the bottom drawer were ruined. I pointed it out to my job leader and asked if they would replace it and guess what he said? Yes they would. Except they didn't. I spent months asking him about it and emailing him about it and even found an equivalent one that was like $300 that they agreed to buy to replace it, but they never ordered it and he would make a big show of having forgotten it and promised to order it that night every single time. Eventually I stopped asking but by then I had nmny breakdown and couldn't afford to replace it. I would have been disappointed but would've accepted if they flat out told me they weren't responsible for the damage, but to yank my chain around for nine months with promises they very carefully avoided ever putting in writing.
Its been years and I'm still angry at that company.
