Capability Study - Making wine bottle corks - Length

[dame]

Hi,
I am studying mechanical engineering at university in the UK.
and I have been given a Quality assurance question but am struggling to get to grips with it. I was wondering if you could help me in anyway either by offering some advice or giving me some contacts who could help. Any assistance would be most grateful.

Thanks, Dame.

Say the making of wine corks to a length of 50mm with data variables say 49.3 50.5 51.2 49.8 etc etc how can I use this data to assess the capability of the machine and consider which specification should be accepted and given out of these:
50mm +-1.5
50mm +-2.5
50mm +0 -6
50mm +-4

 

[D.Scott]

Dame - Welcome to the Cove.

Sounds like you are looking to do a reverse capability. Assuming you know how to determine the standard deviation, a good estimate of needed tollerance can be obtained by taking the average of your process and adding/subtracting the required number of standard deviation.

If you need a final capability of 1.00, you would multiply the S/D by 3 and add/subtract to/from the average. 1.33 = x4; 1.67 = x5; 2.00 = x6.

You cannot assess the machine's capability until you have a tollerance to compare with.

I hope this is on the right line for you - if I have misunderstood, please re-post for more info.

Any others who can offer other suggestions, please jump right in.

Dave

 

[Rick Goodson]

Dame,

IMHO Dave is certainly correct. I would caution however that when and how the data was collected would influence your choice of the index. If the data was al gathered at one point in time from a continuous process we may safely assume that the sample mean will move over time. Selecting an index of 1.0 would not allow for that movement. I recommend 1.67 as the index for your calculation.

Regards,

Rick

 

[KenK - 2009]

Of course the real answer is "YOU"RE DOING THIS BACKWARDS!!"

as hinted to in a previous reply.

What you are supposed to do is research the effect of the cork dimension with respect to its impact on customer satisfaction using simulations and testing (DOEs!). This could include wine quality, wine life, propensity for the bottle neck to crack, damage rate to corks during manufacturing, effort needed to remove the cork, etc.... You could even consider other possible cork materials that might provide better results.

All of these potential impacts would need to be assessed w.r.t. cork dimension and then summarized to determine the appropriate specification limits for the cork.

If the cork is too tight, the customer may not cleanly remove it, plus other manufacturing and life issues could arrise. If the cork is too loose, then the quality of the wine could suffer, not to mention the economics of spillage.

You don't determine what's acceptable based on what's available on hand - you determine it based upon the impact on the product/customer.