Can I still use control chart to control my process even if my process is not capable

[philiplim]

Hi,

My current processs in in statistical control but it's not capable of meeting the specification of 0.63 +0.02 -0.03. My process has done everything to improve and this is the best they can achieve. My question is can I still use control chart to control my process to reduce the scrap level? How shall I establish my control limits so that I can input it in Minitab under the S limits for Upper & Lower bound?

I have done control charts on the 4 IDs and the control limits are covering my spec. limits.

Please help.

 

[Dr. Electron]

Re: Can I still use control chart to control my process even if my process is not cap

Hi Philip,

Before proceeding...there are a few questions that need to be answered:

1. How are you measuring this part? The measuring variability could be contributing a large factor to the variability of the process.

2. What are your control chart and long term Cps and Cpks? (Let me know if you need help calculating these)

3. What actions are you making because of your control charts?

 

[BradM]

Re: Can I still use control chart to control my process even if my process is not cap

Philip,

What is driving the ultra-tight specification? Can it be altered? Have you discussed this with your customer? Are other competitors easily able to meet this specification?

Sometimes, there just isn't much more you can do. And.. what you can do, may be such an expense, it might not be worth it.

As to keeping the control chart, I would always keep it. You will want to know if variance enters your process.

As to reducing scrap level, you may consider other thought-processes in addition to the control chart to reducing waste.

 

[Kales Veggie]

Re: Can I still use control chart to control my process even if my process is not cap

Good point made above.

As part of your process design, you have to determine what your gage variability is. Based on your data, your process is producting close to a CPK and CP of 1.

Yes, you can use control charts for your process, however to prevent shipping NG parts to your customer, you might have to start 100% inspection, especially when the ID is a significant or key characteristic of the product.

My approach:

1) determine your gage variation (if higher than 10-20%, improve your gaging method)
2) determine variation in your process (incoming product, tooling robustness, repeatability of movements in the press, is there any play?).

 

[Geoff Withnell]

Re: Can I still use control chart to control my process even if my process is not cap

Hi,

My current processs in in statistical control but it's not capable of meeting the specification of 0.63 +0.02 -0.03. My process has done everything to improve and this is the best they can achieve. My question is can I still use control chart to control my process to reduce the scrap level? How shall I establish my control limits so that I can input it in Minitab under the S limits for Upper & Lower bound?

I have done control charts on the 4 IDs and the control limits are covering my spec. limits.

Please help.

Not only can you use control charts, I would strongly recommend it. You certainly don't want any assignable cause variation contributing to the already excessive variation you have. Also, the control chart may allow you to see problems with the process yolu might not otherwise catch. In addition to the valuable suggestions put forward by my learned colleagues, I have a suggestion. There appears to be some possibility that you have an over control situation occuring. While there is no clear signal there seems to be a number of instances where an excursion either up or down is following by movement to the other side of the mean, or a period of several up aqnd down sifts before briefly settling down again. If the operator is making adjustments, try a trial run either with no adjustments unless the process goes out of control (NOT out of spec) or noting the time, magnitude and direction of any adjustment or tool change.

Geoff Withnell

 

[Bev D]

Re: Can I still use control chart to control my process even if my process is not cap

Hi,

My current processs in in statistical control but it's not capable of meeting the specification of 0.63 +0.02 -0.03. My process has done everything to improve and this is the best they can achieve.

As previous posters have said - yes you can stilll use control charts to keep the process from getting worse and it might give you clues as to the caus(es) of unwanted variation

However, I would also add that I typically have great skepticism when people tell me that a process can't be made better. I have solved inumerable problems that couldn't be solved. You might also want to start investigating good problem solving techniques. I would recommend "Statistical Engineering" by Steiner and Mackay as a great start...

 

[potdar]

Re: Can I still use control chart to control my process even if my process is not cap

Hi,

My current processs in in statistical control but it's not capable of meeting the specification of 0.63 +0.02 -0.03. My process has done everything to improve and this is the best they can achieve. My question is can I still use control chart to control my process to reduce the scrap level? How shall I establish my control limits so that I can input it in Minitab under the S limits for Upper & Lower bound?

I have done control charts on the 4 IDs and the control limits are covering my spec. limits.

Please help.

I dont use minitab so I cant tell you much about it. I will only say that if you use control charts, simply forget about the specification limits. Use your process control limits to guide you. Take corrective actions only when the control charts show any possible out of limit situation / trend w.r.t. the control limits - not the specification.

What is to be done so that the customer gets within specification parts can be seen later - 100% inspection, what else!

 

[Eric ng]

Re: Can I still use control chart to control my process even if my process is not cap

It depends at what stage of process development. If it is a new process, you should perform a thorough process capability study to ensure the process is stable and the products meet specification limits. Depending on customer's requirement, generally, you can start production if meet a minimum Cpk value of 1.33. Other customer may required a Cpk of 1.67.

If you are trying to analyze an existing process which have not been subjected to a formal capability study before, and the process shows instability, then you can use this information to detemine the root cause(s), and find solutions to eliminate them, to achieve process stability. Until you achieve stability, you should have a containment measure to ensure no out-of-spec products reach customers.

Strictly speaking, it's only meaningful to implement process control using SPC for a stable process, wheter the products meet specifications or not. You can have an unstable process, and yet all products meet specification limits, or a stable process, but some products are outside specificatoin limits.

Eric

 

[Miner]

Re: Can I still use control chart to control my process even if my process is not cap

Hi,

My current processs in in statistical control but it's not capable of meeting the specification of 0.63 +0.02 -0.03. My process has done everything to improve and this is the best they can achieve. My question is can I still use control chart to control my process to reduce the scrap level? How shall I establish my control limits so that I can input it in Minitab under the S limits for Upper & Lower bound?

I have done control charts on the 4 IDs and the control limits are covering my spec. limits.

Please help.

I strongly recommend performing an MSA on the measurement system. Stamped ID holes typically have a breakaway zone of larger diameter than the shear zone. Throw in potential burrs and you may have a lot of measurement variation induced by the part itself. Do you have a standardized measurement method to ensure all operators measure at the same location?

 

[philiplim]

Re: Can I still use control chart to control my process even if my process is not cap

Hi Everyone,

Sorry for the late response. Been trying to solve a quality issue at my side.

We are using profile projector to measure the parts.

Just want to check if anyone of you tried SPCXL before. I downloaded the trial version and try using it. It's quite useful as I am able to key in the number of standard deviations to control chart my process.

I intend to use this software to calculate the control limits at 2.5 sigma and control my process at 2.5 sigma for each ID. At least, this can help to monitor the process and at the same time trigger the process personnel to adjust the tooling if any trend is observed. I am using only 2 trend rules. They are 1) exceeds control limit & 2) Zone A 2 of 3. I believe it can also help me to reduce my scrap level.

2.5 sigma - UCL = 0.647137 & LCL = 0.612129
3.0 sigma - UCL = 0.650638 & LCL = 0.608629

This software is able to draw the control chart for me instead of manual plotting.

We can't open up the spec. as there is quality issue before due out of spec. At this point, I don't think we can ask the customer to change design.

Any comments on the above method. Please feel free to comment.

Regards.