The prcess capability is the capability of the process - ie when the process is running under a constant sysetm of random causes the total variation that is expected from it.
The best and easiest method to be followed is the control chart methodology- data collect, homogenise, and using the statistical constant convert the avge range into process sigma.
The machine capability- as the name explains is the machine variance and the other 5Ms of the process are now taken out. This is just a bit difficult ball game.
here we do not work under constant sustem of Chance cause but under constant chance cause. Usually it requires a mass production machinery.
Here you have to collect at least 50 consecutive readings from an unterrupted run and calculate their std dev.
In acse any parameter is changed or the sequence is broken for whatsoever reason, start back.
Each has its own place in process improvement and process analysis.
Hope it clarifies.
__________________
Fantastic ideas for a Fantastic World, I make the illogical logical.They put me in cars,they put me in yer tv.They put me in stereos and those little radios you stick in your ears.They even put me in watches, they have teeny gremlins for your watches
If you like i can send you a machine capability sheet I have in excell that i use. You just fill in the blanks and it will give you the efficiency of the machine it is very easy.
One of the best software packs as I have used - if you are in the quality business is minitab- most of the things you want to do in normal course of business can be done by it.
However it is not freeware and you have to purchase it.
SL-
The process capability is just a bit different- as I said in the post#1, you need to homogenise data. For thsi purpose, it may be just a bit difficault, and a few look-up tables will be required and at least a few macros to run.
Never tried it in the excell, but feel may be just a bit too complicated (infact didn't try writing program in excell for machine capabilities too).
__________________
Fantastic ideas for a Fantastic World, I make the illogical logical.They put me in cars,they put me in yer tv.They put me in stereos and those little radios you stick in your ears.They even put me in watches, they have teeny gremlins for your watches
I dont see where i can insert a file on here. But if you send me your email I will gladly forward. U must have excel to use it. It is basically where you fill in your expected machines target output. say 500 per day you enter the real number that you are getting. The time in the day in minutes say 8 hrs = 480 mins. then downtime for repair or maintainance. it calculates everything for you. You just have to input the information. Just let me know if you would like it. I use it everyday in manufacturing. It gives you a good idea if your machine is operation as expected. Or if you have an operator. and there are no break downs during the day. You can then look at why you are not producing at 100% machine or operator.
The machine capability is tha capability of the machine to hold a given tolerance (not exact meaning but conceptually) It is just a bit different from our accuracy, repeatibility etc that are being done during machine calibration / accuracy checks.
This is the band that is achieved when the machine is on the operation. You might have seen a number of times that when a perfectly calibrated machine is installed witth accuracy 20 microns/m and total max positional error of 35 microns /m your variation in actual piece is much larger than this. How much is studied by the machine capability.
We argue that it might be due to different operator , material etc. When those are taken in account (They along with machine, are parts of the 6M that forms the process)
However if we try to keep all other things constant, then what machine achieves is the machine capability
__________________
Fantastic ideas for a Fantastic World, I make the illogical logical.They put me in cars,they put me in yer tv.They put me in stereos and those little radios you stick in your ears.They even put me in watches, they have teeny gremlins for your watches
Also if you are looking to do a process capability to get the cpk or ppk of a run. Its all about the data. I use a histogram when running our process. it stays up on the screen and the operators input there data....whether it be length or attributes.....it all shows how well your process is holding , If you average is running to the high side of the process you should be alerted to make an adjustment to get it closer to your target......mostly done with macros......collect the data......say 5 samples say your target is 500 ...your samples you take are 501,502,500,509,506 now add them up. 2518.divide by the numbers of samples you have just taken 5 samples divided by 2518 =503 that is your average so you are running 3 off your target on average...depending what your tolerance is. It would show if you have a good process capability.
When I read your question it truely inspired some internal thought... I love that. So I set about identifying different perspectives, and then i stumbled across this link...
I believe the information therein is the answer you are looking for... at least for a "standards" perspective. Meanwhile, I will keep the debate going among the demons within me...
"Almost" Good Answers: