User Panel
User Panel
User Panel
|
Posted: 1/14/2023 12:15:20 AM EDT
[#1]
Quote HistoryQuoted: There were 2 scrap. Running 250 parts, I get the process down to predictable by 100% inspection of about the first 20. Then every 10 or so parts for critical dims or more if there is anything wonky. Like if that inspected part gets close to the edge, I'll look closely at the tool and setup to determine why. Its kind of an "on the fly" application of what the cpk end result shows. Only I can catch the issue before running 250 parts, instead of looking at some computer data after they're all ran and saying " Dang, these are all over the place" or "Awesome job, ridgerunner precision, as usual".  In mass production those 2 bad per 250 is terrible, I cover about 100 million components inbound to our plant a year, we aim for a PPM (defects per million) of under 50-this cast, stamped, machine, plastic, rubber etc. Your Cpk can help you understand process variations and help target those issues for elimination. Setting control limits at 75% of the tolerance is a good start to reducing defects. Not trying to insult you but the 2 bad parts is 8000 PPM, we would be in weekly meetings and doing 100% inspection on every part with that performance if I had a supplier shipping at those rates. |
|
|
|
Posted: 1/14/2023 12:18:30 AM EDT
[#2]
This kind of stuff makes me glad that I'm doing one-up repair jobs, instead of production. I do have to work to rather close tolerances where I have to worry about ambient temperature. But, if I screw-up, I can just re-do the work. Though, that becomes a short term profit loss for the company.
|
|
|
|
Posted: 1/14/2023 12:21:23 AM EDT
[#3]
Quote HistoryQuoted: Cpk is a way of expressing the capability of a process to meet a customer specification. Process capability is generally reviewed over multiple batches/data points to get an understanding of the stability of the outputs but that gets into Cpk vs Ppk. A work station may make all parts within spec, but how tight in relation to the tolerance, is it centered within the tolerance range, is it skewed to one side? These can be indicators of long term performance. https://www.advanceinnovationgroup.com/blog/uploads/images/image_750x_5cc18f061482a.jpg Here’s to hoping your end product’s distribution is mesokurtic! |
|
|
|
Posted: 1/14/2023 12:25:34 AM EDT
[#4]
Quote HistoryQuoted: In mass production those 2 bad per 250 is terrible, I cover about 100 million components inbound to our plant a year, we aim for a PPM (defects per million) of under 50-this cast, stamped, machine, plastic, rubber etc. Your Cpk can help you understand process variations and help target those issues for elimination. Setting control limits at 75% of the tolerance is a good start to reducing defects. Not trying to insult you but the 2 bad parts is 8000 PPM, we would be in weekly meetings and doing 100% inspection on every part with that performance if I had a supplier shipping at those rates. Quote HistoryQuoted: Quoted: There were 2 scrap. Running 250 parts, I get the process down to predictable by 100% inspection of about the first 20. Then every 10 or so parts for critical dims or more if there is anything wonky. Like if that inspected part gets close to the edge, I'll look closely at the tool and setup to determine why. Its kind of an "on the fly" application of what the cpk end result shows. Only I can catch the issue before running 250 parts, instead of looking at some computer data after they're all ran and saying " Dang, these are all over the place" or "Awesome job, ridgerunner precision, as usual". In mass production those 2 bad per 250 is terrible, I cover about 100 million components inbound to our plant a year, we aim for a PPM (defects per million) of under 50-this cast, stamped, machine, plastic, rubber etc. Your Cpk can help you understand process variations and help target those issues for elimination. Setting control limits at 75% of the tolerance is a good start to reducing defects. Not trying to insult you but the 2 bad parts is 8000 PPM, we would be in weekly meetings and doing 100% inspection on every part with that performance if I had a supplier shipping at those rates. Mass production vs custom shop. His customer is wanting to ramp up and is trying to apply the small batch shop quality to a factory setting. It won’t work, but sounds good to his bosses. |
|
|
|
Posted: 1/14/2023 4:28:30 AM EDT
[#5]
Quote HistoryQuoted: In mass production those 2 bad per 250 is terrible, I cover about 100 million components inbound to our plant a year, we aim for a PPM (defects per million) of under 50-this cast, stamped, machine, plastic, rubber etc. Your Cpk can help you understand process variations and help target those issues for elimination. Setting control limits at 75% of the tolerance is a good start to reducing defects. Not trying to insult you but the 2 bad parts is 8000 PPM, we would be in weekly meetings and doing 100% inspection on every part with that performance if I had a supplier shipping at those rates. He’s prototyping, there’s gonna be a couple bad parts and they paid for those too. |
|
|
|
Posted: 1/14/2023 6:03:49 AM EDT
[#6]
Quote HistoryQuoted: In mass production those 2 bad per 250 is terrible, I cover about 100 million components inbound to our plant a year, we aim for a PPM (defects per million) of under 50-this cast, stamped, machine, plastic, rubber etc. Your Cpk can help you understand process variations and help target those issues for elimination. Setting control limits at 75% of the tolerance is a good start to reducing defects. Not trying to insult you but the 2 bad parts is 8000 PPM, we would be in weekly meetings and doing 100% inspection on every part with that performance if I had a supplier shipping at those rates. Quote HistoryQuoted: Quoted: There were 2 scrap. Running 250 parts, I get the process down to predictable by 100% inspection of about the first 20. Then every 10 or so parts for critical dims or more if there is anything wonky. Like if that inspected part gets close to the edge, I'll look closely at the tool and setup to determine why. Its kind of an "on the fly" application of what the cpk end result shows. Only I can catch the issue before running 250 parts, instead of looking at some computer data after they're all ran and saying " Dang, these are all over the place" or "Awesome job, ridgerunner precision, as usual". In mass production those 2 bad per 250 is terrible, I cover about 100 million components inbound to our plant a year, we aim for a PPM (defects per million) of under 50-this cast, stamped, machine, plastic, rubber etc. Your Cpk can help you understand process variations and help target those issues for elimination. Setting control limits at 75% of the tolerance is a good start to reducing defects. Not trying to insult you but the 2 bad parts is 8000 PPM, we would be in weekly meetings and doing 100% inspection on every part with that performance if I had a supplier shipping at those rates. LOL. Those two were in the first 4 parts and weren't shipped, obviously. You don't do setup parts? |
|
|
|
Posted: 1/14/2023 6:22:28 AM EDT
[#7]
Ask ChatGPT.
|
|
|
|
Posted: 1/14/2023 6:28:58 AM EDT
[#8]
Quoted: I been machining and designing a long time. Sometimes I get in pretty deep but have always felt comfortable that I'll figure it out. No problem. I always have. Even with some pretty complex or difficult tasks. I'm sure this will be simple for math guys or, specifically, production engineers. But I got asked my opinion on a specific type of "tolerancing" that I'm not familiar with. I made 250 parts for a customer. They were in tolerance. Once specific tolerance of one feature was +/- .03mm. Customer writes: "These XXXXXX turned out great – awesome job. Per the usual, not surprising ?? (eta: the smiley face posted as question marks) I failed to relay a capability request to you – the Design Engr was hoping for 1.33cpk min on this 5.47 +/- .03mm dim. You being the wizard you are, hit 1.33 cpk without even babysitting it (data below). Anyway, long-term, I hope to MIM this part. If I get jammed up for some reason and need to machine this part long term, there’s no reason any shop worth its salt couldn’t hit 1.33pk on this dim/tol, right? Just a single tool making a pass, and as long as wear and radil comp is monitored…I would think it wouldn’t be an issue. Your thoughts? " I never heard of cpk.  So I looked it up. Fuck.  https://sixsigmastudyguide.com/process-capability-cp-cpk/#:~:text=How%20to%20Calculate%20Cpk,-Cpk%20is%20a%20measure%20to When I have to start looking up the terms they're using to get the shit they need like Z-score, Standard Deviation, Specification limit....and all the fargin' formulas to find them. My head hurts. I'll be able to get it eventually but this shit is 'tarded. Egg head. College whizz kid shit. I understand the value of the info for production and prediction of processes. But, I had to tell him that if he asks the average shop to be able to hit a 1.33cpk, I'm not sure how many are going to understand the request. First time in my career I've felt a little under equipped.  you are right where you need to be. never stop learning. "fake it until you make it" is a viable road to success. |
|
|
|
Posted: 1/14/2023 6:30:01 AM EDT
[#9]
I feel ya OP .
Some of the shit they’ve got in surveying now blows my mind . A scanner hat takes millions upon millions of shots , drone surveying . That’s one thing I should do this winter is get my drone flying license . I don’t really want to do that shit , but gotta stay useful to the company . |
|
|
|
Posted: 1/14/2023 6:33:39 AM EDT
[#10]
Quote HistoryQuoted: Tolerances are tolerances.  Sounds like the latest corporate double speak to make someone sound smart.My thoughts, but there is more to it than that. Sounds to me like they're trying to outsource their QC and specs to the manufacturer, which is fine, but they're trying to let you do the calculations to meet the metric of their process instead of doing it themselves. "Listen smarty pants, I produced in-spec parts, you want to calculate cpk and your CSU and CSL it's up to you to sit down and do it, while you manufacture in-house. Otherwise quit trying to outsource this shit and pay up to get your parts." Are you doing all this on CNC? See if its possible that they hire you to run an inhouse machine shop for them, basically making them your only client, if that's something you'd want. Calculate what you could save them over time and make an offer. My opinion at least and it's what I did once to work myself into a high paying job with full bennies. |
|
|
|
Posted: 1/14/2023 6:34:45 AM EDT
[#11]
Tell him to delete the newbie's copy of Minitab and get on with life.
|
|
|
|
Posted: 1/14/2023 7:04:23 AM EDT
[#12]
Quote HistoryQuoted: LOL. I have pretty good people skills. These guys love me. I've saved their ass many times. But this bit threw me and I didn't want to give an uneducated answer. Trying to get a grasp on the concept in a couple hours was....it hurt my wittle head.  Quote HistoryQuoted: Quoted: basically it sounds like you need to hire a guy with 'people skills' to act as your go-between https://www.youtube.com/watch?v=hNuu9CpdjIo just a funny clip to cheer ya up OP  LOL. I have pretty good people skills. These guys love me. I've saved their ass many times. But this bit threw me and I didn't want to give an uneducated answer. Trying to get a grasp on the concept in a couple hours was....it hurt my wittle head. Maybe playing with some real numbers will help? Put the finished dimensions from the parts you made in here. That'll give you the mean and standard deviation. https://www.calculator.net/standard-deviation-calculator.html Then put those numbers plus the upper and lower acceptable dimensions (from the tolerances) in here and it'll give you the CPK. https://calculator.academy/cpk-calculator/#f1p1 CP just requires a small standard deviation i.e. a tight group. They could all be at the small end of the tolerance and that would be fine. CPK requires a small standard deviation and an average size that's close to the middle of the tolerance range. Once you've done that with your actual dimensions, try adjusting your numbers - for example, make all the parts 0.01 smaller. That's the same range and distribution but less well centered. Have a look at how that changes your CPK. Throw in a few outliers that are barely within tolerance and see the effect. |
|
|
|
Posted: 1/14/2023 9:08:31 AM EDT
[#13]
It's a fancy way of expressing a simple concept.
The basic idea is that they want to make sure that your manufacturing process is under control. If you're at your maximum tolerance for one piece and at the minimum for the next, it means that you have an increased likelihood of going outside the allowable tolerance range. They don't really care if you're on the high side or the low side (perfect would be better  ) as long as your variation is controlled. Wide swings are bad. |
|
|
|
Posted: 1/14/2023 9:24:59 AM EDT
[#14]
Quoted: I been machining and designing a long time. Sometimes I get in pretty deep but have always felt comfortable that I'll figure it out. No problem. I always have. Even with some pretty complex or difficult tasks. I'm sure this will be simple for math guys or, specifically, production engineers. But I got asked my opinion on a specific type of "tolerancing" that I'm not familiar with. I made 250 parts for a customer. They were in tolerance. Once specific tolerance of one feature was +/- .03mm. Customer writes: "These XXXXXX turned out great – awesome job. Per the usual, not surprising ?? (eta: the smiley face posted as question marks) I failed to relay a capability request to you – the Design Engr was hoping for 1.33cpk min on this 5.47 +/- .03mm dim. You being the wizard you are, hit 1.33 cpk without even babysitting it (data below). Anyway, long-term, I hope to MIM this part. If I get jammed up for some reason and need to machine this part long term, there’s no reason any shop worth its salt couldn’t hit 1.33pk on this dim/tol, right? Just a single tool making a pass, and as long as wear and radil comp is monitored…I would think it wouldn’t be an issue. Your thoughts? " I never heard of cpk. So I looked it up. Fuck. https://sixsigmastudyguide.com/process-capability-cp-cpk/#:~:text=How%20to%20Calculate%20Cpk,-Cpk%20is%20a%20measure%20to When I have to start looking up the terms they're using to get the shit they need like Z-score, Standard Deviation, Specification limit....and all the fargin' formulas to find them. My head hurts. I'll be able to get it eventually but this shit is 'tarded. Egg head. College whizz kid shit. I understand the value of the info for production and prediction of processes. But, I had to tell him that if he asks the average shop to be able to hit a 1.33cpk, I'm not sure how many are going to understand the request. First time in my career I've felt a little under equipped. Did they have you design the process or just make the prototype parts? CPK is on the process for production. That’ll be the production shop’s problem at the end of the day. ETA: re-read his question. Too many variables in machines, fixturing, and talent to answer that. Tell him the answer is “yes, as long as they process it correctly on a decent machine.” |
|
|
|
Posted: 1/14/2023 9:29:29 AM EDT
[#15]
When customers start talking like that, we turn them away. Lol.
You just know they will be a problem long term. |
|
|
|
Posted: 1/14/2023 9:37:59 AM EDT
[#16]
Where I work, we had an electronic charting system that was rock solid and simple to use. I used it from 2003 to 2021. Then they decided to scrap the entire thing and roll out a new EMR. They trained us for a week in a class. There were several online classes. I was more confused after the training. The color scheme is hard for me to see the text. Thin grey text on a blue background, fucking stupid. The system is glitchy on a good day. It freezes up alot, boots users out and doesn't save what we charted sometimes. I am struggling to learn this system. There is no easy way to chart and you have to go through several menu options to enter a simple note. The system relies on a scanner to chart meds and labs. The damn scanner won't recognize meds or the patient often. It's a nightmare to use and has added double the time to chart than the previous EMR.
Personally, I've gone from 100% charting reviews down to less than 50%. Most of the errors are from the system not saving what was entered. Most everyone is fed up. Four years left of that bullshit. |
|
|
|
Posted: 1/14/2023 9:46:52 AM EDT
[#17]
I just want to say on behalf of we mathematically retarded knuckle draggers of the world that I am glad you folks exist. Thank you.
|
|
|
|
Posted: 1/14/2023 9:48:08 AM EDT
[#18]
Quote HistoryQuoted: you are right where you need to be. never stop learning. "fake it until you make it" is a viable road to success. I've faked it pretty well so far, since 1986. I hope nobody catches on.
|
|
|
|
Posted: 1/14/2023 9:48:12 AM EDT
[#19]
Quote HistoryQuoted: In mass production those 2 bad per 250 is terrible, I cover about 100 million components inbound to our plant a year, we aim for a PPM (defects per million) of under 50-this cast, stamped, machine, plastic, rubber etc. Your Cpk can help you understand process variations and help target those issues for elimination. Setting control limits at 75% of the tolerance is a good start to reducing defects. Not trying to insult you but the 2 bad parts is 8000 PPM, we would be in weekly meetings and doing 100% inspection on every part with that performance if I had a supplier shipping at those rates. Based on previous conversations I think he did these on a manual machine...you are also checking inbound components, which means the ones you see are the ones that made it though some other operations quality process. Honestly you should never see any defects, you're the customer. What I was alluding to was what we call setup parts. Typically in machining, when you are making a volume of parts you'll have a setup part or two. Of course you can make the first part perfect but it's a different method and depending on the equipment, sneaking up on a dimension doesn't guarantee that the next part will be in tolerance. I've done both, millions of parts a year for 20 years to a company that built outdoor lighting and one piece high precision hardened tool steel components for too and die work. Your comparison is invalid. They are different worlds. |
|
|
|
Posted: 1/14/2023 9:51:06 AM EDT
[#20]
Quote HistoryQuoted: There were 2 scrap. Running 250 parts, I get the process down to predictable by 100% inspection of about the first 20. Then every 10 or so parts for critical dims or more if there is anything wonky. Like if that inspected part gets close to the edge, I'll look closely at the tool and setup to determine why. Its kind of an "on the fly" application of what the cpk end result shows. Only I can catch the issue before running 250 parts, instead of looking at some computer data after they're all ran and saying " Dang, these are all over the place" or "Awesome job, ridgerunner precision, as usual". Been a minute since I did real work, but what you're describing is probably what they're asking. Sounds like they checked them all, did the math, and you came out fine. What they want to know from you (I think) is how hard is it to have a process that keeps them as consistent as yours were? So the things you were/are checking in the setup/tooling when the dimension begins to wander. The reason they're asking (obviously) is that they'd like to MIM them, but don't know if MIM will hold the dimension at that capability, so they're trying to figure what the rework would be if needed. |
|
|
|
Posted: 1/14/2023 9:52:08 AM EDT
[#21]
I'm an Automation Engineer and one of the things I say to people all the time is "speak English and tell me what you want in 15 words or less". Usually they will think for about 15 seconds and tell you exactly what they want.
There are two types of engineers, those that want to make a better X and those that want to prove how smart they are. Those that are out to prove how smart they are design things that work most of the time but the user has no idea how it works. The first type of engineer makes simple things that all the operators understand. Nobody needs a radio with 60 knobs. |
|
|
|
Posted: 1/14/2023 9:52:24 AM EDT
[#22]
Customers like a higher Cpk value so they get a good product at the lowest cost. Remember it is all about the dollar.
Higher values mean that the likelihood of you producing and out of spec part are lower so you have to spend less money verifying dimensions, adjusting your process, scrapping those that are out, etc. All of these translate to lower costs for you and subsequently the customer. ie if you have a higher cpk, your process should be able to produce the product cheaper than your competitors with lower cpk values. and yes, I hate it when incompetent engineers get a copy of minitab and bs their way through not understanding the physics with statistical hand waving. |
|
|
|
Posted: 1/14/2023 9:55:01 AM EDT
[#23]
Quote HistoryQuoted: Maybe playing with some real numbers will help? Put the finished dimensions from the parts you made in here. That'll give you the mean and standard deviation. https://www.calculator.net/standard-deviation-calculator.html Then put those numbers plus the upper and lower acceptable dimensions (from the tolerances) in here and it'll give you the CPK. https://calculator.academy/cpk-calculator/#f1p1 CP just requires a small standard deviation i.e. a tight group. They could all be at the small end of the tolerance and that would be fine. CPK requires a small standard deviation and an average size that's close to the middle of the tolerance range. Once you've done that with your actual dimensions, try adjusting your numbers - for example, make all the parts 0.01 smaller. That's the same range and distribution but less well centered. Have a look at how that changes your CPK. Throw in a few outliers that are barely within tolerance and see the effect. Quote HistoryQuoted: Quoted: Quoted: basically it sounds like you need to hire a guy with 'people skills' to act as your go-between https://www.youtube.com/watch?v=hNuu9CpdjIo just a funny clip to cheer ya up OP LOL. I have pretty good people skills. These guys love me. I've saved their ass many times. But this bit threw me and I didn't want to give an uneducated answer. Trying to get a grasp on the concept in a couple hours was....it hurt my wittle head. Maybe playing with some real numbers will help? Put the finished dimensions from the parts you made in here. That'll give you the mean and standard deviation. https://www.calculator.net/standard-deviation-calculator.html Then put those numbers plus the upper and lower acceptable dimensions (from the tolerances) in here and it'll give you the CPK. https://calculator.academy/cpk-calculator/#f1p1 CP just requires a small standard deviation i.e. a tight group. They could all be at the small end of the tolerance and that would be fine. CPK requires a small standard deviation and an average size that's close to the middle of the tolerance range. Once you've done that with your actual dimensions, try adjusting your numbers - for example, make all the parts 0.01 smaller. That's the same range and distribution but less well centered. Have a look at how that changes your CPK. Throw in a few outliers that are barely within tolerance and see the effect. Thanks. That's pretty much what I did last night. I'm a "I need to see it in action" sort. Funny, I was sitting here this morning discussing it with my boy who is about to graduate a MechE. Little shit started spouting off, "Oh, yeah, you take blah blah blah square, divide, blah blah, and then blah blah which gives you blah blah, so you can know blah blah blah." So now I understand it perfectly and will never use it. 
|
|
|
|
Posted: 1/14/2023 10:02:24 AM EDT
[#24]
Quote HistoryQuoted: Cpk of 1.33 isn't that difficult to hit, but obviously the tighter the tolerance, the more difficult it is to hold it in the middle. Been a minute since I did real work, but what you're describing is probably what they're asking. Sounds like they checked them all, did the math, and you came out fine. What they want to know from you (I think) is how hard is it to have a process that keeps them as consistent as yours were? So the things you were/are checking in the setup/tooling when the dimension begins to wander. The reason they're asking (obviously) is that they'd like to MIM them, but don't know if MIM will hold the dimension at that capability, so they're trying to figure what the rework would be if needed. I don't think MIM is going to cut the mustard. I didn't even want to get in to GD&T with him because this dimension is from the centerline of a hole to a surface parallel with the hole's axis. So, he really should take into consideration that the tolerance on the hole is +/- .038.
|
|
|
|
Posted: 1/14/2023 10:04:12 AM EDT
[#25]
You should get your work to pay for a class on SS and all that.
No telling the amount of benefit they'd receive having someone with actual hands-on experience getting some training on this. |
|
|
|
Posted: 1/14/2023 10:10:40 AM EDT
[#26]
Quote HistoryQuoted: I don't think MIM is going to cut the mustard. I didn't even want to get in to GD&T with him because this dimension is from the centerline of a hole to a surface parallel with the hole's axis. So, he really should take into consideration that the tolerance on the hole is +/- .038. |
|
|
|
Posted: 1/14/2023 10:16:44 AM EDT
[#27]
Just make good parts to their tolerances and you will be fine.
|
|
|
|
Posted: 1/14/2023 10:17:05 AM EDT
[#28]
|
|
|
|
Posted: 1/14/2023 10:18:33 AM EDT
[#29]
I stopped caring about six sigma and all that when I realized that it was a way of boiling common sense down to math problems for engineers that were too lazy to get out in the shop and understand the manufacturing needed for their parts.
|
|
|
|
Posted: 1/14/2023 10:22:09 AM EDT
[#30]
Quote HistoryQuoted: You should get your work to pay for a class on SS and all that. No telling the amount of benefit they'd receive having someone with actual hands-on experience getting some training on this. I asked the boss. He laughed and said, "Ain't nobody got time fo dat". |
|
|
|
Posted: 1/14/2023 10:26:25 AM EDT
[#31]
Quote HistoryQuoted: I stopped caring about six sigma and all that when I realized that it was a way of boiling common sense down to math problems for engineers that were too lazy to get out in the shop and understand the manufacturing needed for their parts. In high volume automotive style production I can see the benefit. Low volume high mix it's much harder to make it pay off. But there are good concept to take away. |
|
|
|
Posted: 1/14/2023 2:56:59 PM EDT
[#32]
Quote HistoryQuoted: It has its uses, but it's not the right tool for every problem. Unfortunately the companies that really buy into it tend to think it's a cure all. In high volume automotive style production I can see the benefit. Low volume high mix it's much harder to make it pay off. But there are good concept to take away. Quote HistoryQuoted: Quoted: I stopped caring about six sigma and all that when I realized that it was a way of boiling common sense down to math problems for engineers that were too lazy to get out in the shop and understand the manufacturing needed for their parts. In high volume automotive style production I can see the benefit. Low volume high mix it's much harder to make it pay off. But there are good concept to take away. Getting managers to understand that change should only happen when there's a specific problem to be solved and that you should measure to ensure that it did actually fix the problem is a great thing. If we could make that standard practice everywhere (and for legislators as well) that would be awesome. Beyond that, meh. |
|
|
|
Posted: 1/14/2023 3:02:21 PM EDT
[#33]
Quote HistoryQuoted: Thanks. That's pretty much what I did last night. I'm a "I need to see it in action" sort. Funny, I was sitting here this morning discussing it with my boy who is about to graduate a MechE. Little shit started spouting off, "Oh, yeah, you take blah blah blah square, divide, blah blah, and then blah blah which gives you blah blah, so you can know blah blah blah." So now I understand it perfectly and will never use it. Quote HistoryQuoted: Quoted: Quoted: Quoted: basically it sounds like you need to hire a guy with 'people skills' to act as your go-between https://www.youtube.com/watch?v=hNuu9CpdjIo just a funny clip to cheer ya up OP LOL. I have pretty good people skills. These guys love me. I've saved their ass many times. But this bit threw me and I didn't want to give an uneducated answer. Trying to get a grasp on the concept in a couple hours was....it hurt my wittle head. Maybe playing with some real numbers will help? Put the finished dimensions from the parts you made in here. That'll give you the mean and standard deviation. https://www.calculator.net/standard-deviation-calculator.html Then put those numbers plus the upper and lower acceptable dimensions (from the tolerances) in here and it'll give you the CPK. https://calculator.academy/cpk-calculator/#f1p1 CP just requires a small standard deviation i.e. a tight group. They could all be at the small end of the tolerance and that would be fine. CPK requires a small standard deviation and an average size that's close to the middle of the tolerance range. Once you've done that with your actual dimensions, try adjusting your numbers - for example, make all the parts 0.01 smaller. That's the same range and distribution but less well centered. Have a look at how that changes your CPK. Throw in a few outliers that are barely within tolerance and see the effect. Thanks. That's pretty much what I did last night. I'm a "I need to see it in action" sort. Funny, I was sitting here this morning discussing it with my boy who is about to graduate a MechE. Little shit started spouting off, "Oh, yeah, you take blah blah blah square, divide, blah blah, and then blah blah which gives you blah blah, so you can know blah blah blah." So now I understand it perfectly and will never use it. Get the boy to make you a spreadsheet that takes the nominal, tolerance and a list of dimensions then generates a little infographic with the CP, CPK and a graph of the distribution. Nice little value-add you can give to customers that care. |
|
|
|
Posted: 1/14/2023 3:05:14 PM EDT
[#34]
Seems like a way to ensure that tolerances over time will be in spec, and therefore can randomize Q.C. to a higher degree.
I know nothing of this stuff. |
|
|
|
Posted: 1/14/2023 3:17:07 PM EDT
[#35]
Quote HistoryQuoted: Having a production unit shut down, waiting on parts because of an “in-spec but different” batch, gets expensive very quick. Blame the person whose name is on the print… < knows enough 6Sigma to be dangerous thanks to AlliedSignal/Honeywell. |
|
|
|
Posted: 1/14/2023 3:34:08 PM EDT
[#36]
Quote HistoryQuoted: Get the boy to make you a spreadsheet that takes the nominal, tolerance and a list of dimensions then generates a little infographic with the CP, CPK and a graph of the distribution. Nice little value-add you can give to customers that care. Quote HistoryQuoted: Quoted: Quoted: Quoted: Quoted: basically it sounds like you need to hire a guy with 'people skills' to act as your go-between https://www.youtube.com/watch?v=hNuu9CpdjIo just a funny clip to cheer ya up OP LOL. I have pretty good people skills. These guys love me. I've saved their ass many times. But this bit threw me and I didn't want to give an uneducated answer. Trying to get a grasp on the concept in a couple hours was....it hurt my wittle head. Maybe playing with some real numbers will help? Put the finished dimensions from the parts you made in here. That'll give you the mean and standard deviation. https://www.calculator.net/standard-deviation-calculator.html Then put those numbers plus the upper and lower acceptable dimensions (from the tolerances) in here and it'll give you the CPK. https://calculator.academy/cpk-calculator/#f1p1 CP just requires a small standard deviation i.e. a tight group. They could all be at the small end of the tolerance and that would be fine. CPK requires a small standard deviation and an average size that's close to the middle of the tolerance range. Once you've done that with your actual dimensions, try adjusting your numbers - for example, make all the parts 0.01 smaller. That's the same range and distribution but less well centered. Have a look at how that changes your CPK. Throw in a few outliers that are barely within tolerance and see the effect. Thanks. That's pretty much what I did last night. I'm a "I need to see it in action" sort. Funny, I was sitting here this morning discussing it with my boy who is about to graduate a MechE. Little shit started spouting off, "Oh, yeah, you take blah blah blah square, divide, blah blah, and then blah blah which gives you blah blah, so you can know blah blah blah." So now I understand it perfectly and will never use it. Get the boy to make you a spreadsheet that takes the nominal, tolerance and a list of dimensions then generates a little infographic with the CP, CPK and a graph of the distribution. Nice little value-add you can give to customers that care. That's actually what my customer sent me from their QC dept. I have no interest in production or related stuff. I make parts that nobody else wants to make, sometimes design and build mechanisms they dont know how or have time for, charge a reasonable premium, and deliver in tolerance, nice looking stuff. It's worked for a long time.  Same customer emailed a print hoping I could help them out. Needed two pieces of a 3/16 x .5 lead, two start, left hand, modified buttress screw 4"long. They only shop to quote was minimum 1000 pieces with 12 week lead at a Swiss turn shop. I did them.  Attached FileAfter a few similar jobs he had occasion to visit the shop for the first time. He was standing there and finally asked (trying to honestly not be condescending ) "So....you did all of the parts for us...one these machines?"   Attached FileIt's the Indian, not the arrow. That's how I get business. |
|
|
|
Posted: 1/14/2023 3:50:11 PM EDT
[#37]
Quote HistoryQuoted: Six Smegma. Run away. No better way to fuck up a process than to let a Green belt or Black belt get involved with it. My Green belt project was projected to avoid $6M in warranty costs in less than 5 years. 6Sigma has some good tools but blindly applying them by people that don’t have a clue about the parts/process can be disastrous. |
|
|
|
Posted: 1/14/2023 3:58:19 PM EDT
[#38]
Former QA guy here, been retired nearly 3 years.
If +/- 3 sigma (standard deviations) is exactly tolerance then your Cpk = 1.00. IIRC, a Cpk value of 1.33 means your normal process variation is using approximately 2/3 or the tolerance band and is pretty well centered on the nominal dimension. |
|
|
|
Posted: 1/14/2023 4:30:20 PM EDT
[#39]
Quote HistoryQuoted: This is a good customer. I will follow up on my original email that basically said, most other shops won't know what your talking about. I will explain what the 1.33 capability does in reality for someone quoting the work. I'd quote it to be half the specified tolerance. Quote HistoryQuoted: Quoted: Or cover their ass... Ridgerunner, you can fire customers , if needed. At this point in your career, I doubt you would miss some unreasonable pains in the ass. This is a good customer. I will follow up on my original email that basically said, most other shops won't know what your talking about. I will explain what the 1.33 capability does in reality for someone quoting the work. I'd quote it to be half the specified tolerance. No doubt. You have to decide if your sense of loyalty justifies the brain bending. I'm not a business man but maybe you can get compensated for your extra time and effort (maybe that's a given). I can relate to the struggling part. At the end of my career I spent several years at a new job that some tasks where over my head. We were promised training for some specialized tasks but it never was given and I had to muddle my way through doing an unsatisfactory job. When questioning the supervisor for clues or direction, he would say "Have you ever worked on that before? Well, now you're an expert".  |
|
|
|
Posted: 1/14/2023 5:31:06 PM EDT
[#40]
Quote HistoryQuoted: LOL. Those two were in the first 4 parts and weren't shipped, obviously. You don't do setup parts? Quote HistoryQuoted: Quoted: Quoted: There were 2 scrap. Running 250 parts, I get the process down to predictable by 100% inspection of about the first 20. Then every 10 or so parts for critical dims or more if there is anything wonky. Like if that inspected part gets close to the edge, I'll look closely at the tool and setup to determine why. Its kind of an "on the fly" application of what the cpk end result shows. Only I can catch the issue before running 250 parts, instead of looking at some computer data after they're all ran and saying " Dang, these are all over the place" or "Awesome job, ridgerunner precision, as usual". In mass production those 2 bad per 250 is terrible, I cover about 100 million components inbound to our plant a year, we aim for a PPM (defects per million) of under 50-this cast, stamped, machine, plastic, rubber etc. Your Cpk can help you understand process variations and help target those issues for elimination. Setting control limits at 75% of the tolerance is a good start to reducing defects. Not trying to insult you but the 2 bad parts is 8000 PPM, we would be in weekly meetings and doing 100% inspection on every part with that performance if I had a supplier shipping at those rates. LOL. Those two were in the first 4 parts and weren't shipped, obviously. You don't do setup parts? My post is to help understand WHY they are asking you about the process capabilities. And it sounds like there will be issues in production from the information provided |
|
|
|
Posted: 1/14/2023 5:33:18 PM EDT
[#41]
Quote HistoryQuoted: In mass production those 2 bad per 250 is terrible, I cover about 100 million components inbound to our plant a year, we aim for a PPM (defects per million) of under 50-this cast, stamped, machine, plastic, rubber etc. Your Cpk can help you understand process variations and help target those issues for elimination. Setting control limits at 75% of the tolerance is a good start to reducing defects. Not trying to insult you but the 2 bad parts is 8000 PPM, we would be in weekly meetings and doing 100% inspection on every part with that performance if I had a supplier shipping at those rates. Only 1x/week? |
|
|
|
Posted: 1/14/2023 5:50:09 PM EDT
[#42]
Quoted: Six Smegma. Run away. No better way to fuck up a process than to let a Green belt or Black belt get involved with it. I have 30 years as an aerospace design engineer. My first question would be: Do I have to certify that each batch meets the 1.33 cpk? Then I toss it over to the QA guys, they talk all that BS (and usually get it wrong). Then they get together with the other company's QA, go out for a long lunch, and come up with the magic solution. All the while, you keep cranking out totally acceptable parts, just as you always have. |
|
|
|
Posted: 1/14/2023 5:58:40 PM EDT
[#43]
Quote HistoryQuoted: My post is to help understand WHY they are asking you about the process capabilities. And it sounds like there will be issues in production from the information provided Quote HistoryQuoted: Quoted: Quoted: Quoted: There were 2 scrap. Running 250 parts, I get the process down to predictable by 100% inspection of about the first 20. Then every 10 or so parts for critical dims or more if there is anything wonky. Like if that inspected part gets close to the edge, I'll look closely at the tool and setup to determine why. Its kind of an "on the fly" application of what the cpk end result shows. Only I can catch the issue before running 250 parts, instead of looking at some computer data after they're all ran and saying " Dang, these are all over the place" or "Awesome job, ridgerunner precision, as usual". In mass production those 2 bad per 250 is terrible, I cover about 100 million components inbound to our plant a year, we aim for a PPM (defects per million) of under 50-this cast, stamped, machine, plastic, rubber etc. Your Cpk can help you understand process variations and help target those issues for elimination. Setting control limits at 75% of the tolerance is a good start to reducing defects. Not trying to insult you but the 2 bad parts is 8000 PPM, we would be in weekly meetings and doing 100% inspection on every part with that performance if I had a supplier shipping at those rates. LOL. Those two were in the first 4 parts and weren't shipped, obviously. You don't do setup parts? My post is to help understand WHY they are asking you about the process capabilities. And it sounds like there will be issues in production from the information provided I'm not trying to be an asshole or snarky. May I ask what you do? The reason I ask is I don't think you understand how a machine is set up, dialed in and then ran. The set up took 4 parts to dial in. Nothing was changed for the next twenty with 100% inspection. Nothing was changed for the next 230 with 10% inspection with zero scrap. |
|
|
|
Posted: 1/14/2023 7:24:51 PM EDT
[#44]
Quote HistoryQuoted: I'm not trying to be an asshole or sparky. May I ask what you do? The reason I ask is I don't think you understand how a machine is set up, dialed in and then ran. The set up took 4 parts to dial in. Nothing was changed for the next twenty with 100% inspection. Nothing was changed for the next 230 with 10% inspection with zero scrap. Quote HistoryQuoted: Quoted: Quoted: Quoted: Quoted: There were 2 scrap. Running 250 parts, I get the process down to predictable by 100% inspection of about the first 20. Then every 10 or so parts for critical dims or more if there is anything wonky. Like if that inspected part gets close to the edge, I'll look closely at the tool and setup to determine why. Its kind of an "on the fly" application of what the cpk end result shows. Only I can catch the issue before running 250 parts, instead of looking at some computer data after they're all ran and saying " Dang, these are all over the place" or "Awesome job, ridgerunner precision, as usual". In mass production those 2 bad per 250 is terrible, I cover about 100 million components inbound to our plant a year, we aim for a PPM (defects per million) of under 50-this cast, stamped, machine, plastic, rubber etc. Your Cpk can help you understand process variations and help target those issues for elimination. Setting control limits at 75% of the tolerance is a good start to reducing defects. Not trying to insult you but the 2 bad parts is 8000 PPM, we would be in weekly meetings and doing 100% inspection on every part with that performance if I had a supplier shipping at those rates. LOL. Those two were in the first 4 parts and weren't shipped, obviously. You don't do setup parts? My post is to help understand WHY they are asking you about the process capabilities. And it sounds like there will be issues in production from the information provided I'm not trying to be an asshole or sparky. May I ask what you do? The reason I ask is I don't think you understand how a machine is set up, dialed in and then ran. The set up took 4 parts to dial in. Nothing was changed for the next twenty with 100% inspection. Nothing was changed for the next 230 with 10% inspection with zero scrap. CpK should not be using the 4 set up pieces since it’s measuring a production process. If you get discreet POs for a quantity of parts instead of a production schedule agreement with forecast you really shouldn’t be concerned with this. When I started in aerospace we made parts that were used on the space station - certain people were pushing 6Sigma when we were making one piece….. |
|
|
|
Posted: 1/14/2023 7:56:27 PM EDT
[#45]
I can see this working for the higher up’s to talk there talk on saving money or not losing it by making the wrong size parts. But then the common man has to set up the machine to make the foumla work which in turn make the parts in the specs needed. I’ve seen the sigma crap come and go at a few big companies. It was the thing to do and be. I was in a similar program not as in depth but, trying to be the same on a smaller scale. I took a class had a book. Never talked about it again. It didn’t really help at all. Just fancy words talking about common operations.
|
|
|
|
Posted: 1/15/2023 2:43:15 PM EDT
[#46]
Quote HistoryQuoted: It has its uses, but it's not the right tool for every problem. Unfortunately the companies that really buy into it tend to think it's a cure all. In high volume automotive style production I can see the benefit. Low volume high mix it's much harder to make it pay off. But there are good concept to take away. Quote HistoryQuoted: Quoted: I stopped caring about six sigma and all that when I realized that it was a way of boiling common sense down to math problems for engineers that were too lazy to get out in the shop and understand the manufacturing needed for their parts. In high volume automotive style production I can see the benefit. Low volume high mix it's much harder to make it pay off. But there are good concept to take away. And this is why I lament the proliferation of people who have migrated from automotive into the comparatively high mix/low volume world of heavy equipment. Don't fucking talk to me about PPM for something with a 50 EAU. Jesus Christ.
|
|
|
|
Posted: 1/15/2023 2:44:47 PM EDT
[#47]
Quote HistoryQuoted: After a few similar jobs he had occasion to visit the shop for the first time. He was standing there and finally asked (trying to honestly not be condescending ) "So....you did all of the parts for us...one these machines?" hahahaha |
|
|
|
Posted: 1/15/2023 11:47:09 PM EDT
[#48]
Quote HistoryQuoted: And this is why I lament the proliferation of people who have migrated from automotive into the comparatively high mix/low volume world of heavy equipment. Don't fucking talk to me about PPM for something with a 50 EAU. Jesus Christ. I’ve dealt with 1 EAU as a supplier to a large yellow company…. And having to run performance tests to show that a replacement part made today has the same performance as a 25 y/o version - suppliers have changed, test facilities have closed & new ones built with modern data acquisition & controls vs manometers on the wall…. |
|
|
|
Posted: 1/16/2023 9:39:29 AM EDT
[#49]
Quote HistoryQuoted: I’ve dealt with 1 EAU as a supplier to a large yellow company…. And having to run performance tests to show that a replacement part made today has the same performance as a 25 y/o version - suppliers have changed, test facilities have closed & new ones built with modern data acquisition & controls vs manometers on the wall…. 
|
|
|
|
Posted: 1/16/2023 9:50:49 AM EDT
[#50]
CPK isn't just about making good parts. It's about statistical analysis of dimensional stability.
Basically he wants to know if other people can make this as well as you do. Which is a bullshit question. |
|
|
Win a FREE Membership!
Sign up for the ARFCOM weekly newsletter and be entered to win a free ARFCOM membership. One new winner* is announced every week!
You will receive an email every Friday morning featuring the latest chatter from the hottest topics, breaking news surrounding legislation, as well as exclusive deals only available to ARFCOM email subscribers.
AR15.COM is the world's largest firearm community and is a gathering place for firearm enthusiasts of all types.
From hunters and military members, to competition shooters and general firearm enthusiasts, we welcome anyone who values and respects the way of the firearm.
Subscribe to our monthly Newsletter to receive firearm news, product discounts from your favorite Industry Partners, and more.
Copyright © 1996-2024 AR15.COM LLC. All Rights Reserved.
Any use of this content without express written consent is prohibited.
AR15.Com reserves the right to overwrite or replace any affiliate, commercial, or monetizable links, posted by users, with our own.
