The Elsmar Cove Wiki More Free Files The Elsmar Cove Forums Discussion Thread Index Post Attachments Listing Failure Modes Services and Solutions to Problems Elsmar cove Forums Main Page Elsmar Cove Home Page
Google
  Web Elsmar.com
*Please be aware that SOME RECENT forum threads may not yet be indexed by Google.

View Full Version : Special Processes - Can someone help us with a definition?


Janie
27th August 1998, 03:30 PM
Its me again, Margaret! Sorry if I seem ignorant, but I am! Can someone help us with a definition of Special Processes? I know what the standard says but that doesn't help much. We use ultra-sonic welders to put inserts into plastic parts...is that a special process? It would be obvious before a end user customer got it if the insert were'nt right because the next step of assembly would be impossible. We also apply paint to plastic parts...is that a special process? It is the most part, decoration only.

Marc
27th August 1998, 04:43 PM
Short Definition:

Special processes are those which cannot be verified after the process without destructive testing.

Summary: A special process is any production or service delivery process that generates outputs that cannot be measured, monitored, or verified until it's too late. It's often too late because deficiencies may not be obvious until after the resulting products have been used or services have been delivered. In order to prevent output deficiencies, these special processes must be validated in order to prove that they can generate planned results.

Also see: Special Processes - Can anyone give me a good definition of what they are? (http://elsmar.com/Forums/showthread.php?t=11429)

Marc
27th August 1998, 04:51 PM
Are you Janie or Margaret?

Painting as a 'special process'
You can check thickness with newer 'tools' but it used to be you would have to take a slice through to the base metal.
Let's say the paint is supposed to last 30 years before fadingand you have an accellerated life test to verify that it does last 30 years before fading. From that aspect the process could be called 'special' as ther is not way to verify it without scarificing the test part.

Welds used to be 'special processes' but with x-ray and such you can now see inside of welds making it not so much a special process as it used to be.

Some bomb fuse assembly is 'special process' because once the thing is assembled you 'assume' nothing shifted or anything - and you can't see inside. You have to verify the 'lot' by setting a few off. Air bag modules are the same type of 'you gotta destroy a few to verify the lot' situations.

Confused yet?

Janie
27th August 1998, 08:11 PM
Sorry about the confusion over the name....I was making a feeble joke about a Ray Stevens song.

Yes, I am confused.

We were "handed" a bunch of Procedures right before our audit last December. We managed to achieve certification in spite of the procedures because the auditor took into account we hadn't had them long and we obviously knew what we were doing as far as production and quality. A group of us are now in the process of trying to rewrite the procedures to fit what we actually do. You can expect a lot of dumb questions from me over the next few weeks.

Anyway, back to special processes.
We were told that painting and sonic welding were to be considered special processes but we weren't told why. We have documentation that the people who do painting receive special training and are "qualified" by that training. We agreed that the technicians who adjust the settings of the welders should be qualified and took care of that. Our question now is, did we make things more complicated then they need to be? Why would these be special processes? There are no specifications for thickness of paint except a visual check for coverage. The sonic welding is simply to put a metal insert in so they'll have somewhere for a screw to attach later or to ensure that two parts are joined.
I may have muddied the waters even more now but it sure is nice to have someplace to ask questions and get an answer. Thanks so much!

Marc
27th August 1998, 11:39 PM
As I pointed out with the paint example above, a process is a special process if there is a requirement which cannot 'normally' be verified. You have to be knowledgable enough to confront the auditor and explain there are no critical characteristics (measurement like paint thickness) or functions being verified which cannot be done by 'normal' means (paint coverage - a visual).

As far as the insert, the 'verification' is the next process step its self (you say "It would be obvious before a end user customer got it if the insert were'nt right because the next step of assembly would be impossible." Your company has made a conscious decision to not inspect at the step. Probably for good reason, such as a lot failure rate.

Breakdown:

--> Where the results of processes cannot be fully verified by
--> subsequent inspection and testing of the product and where,

Key word in bold.

--> for example, processing deficiencies may become apparent only
--> after the product is in use,

Are your returns showing any failure of the paint or the inserts? The question becomes, how far do you go. Always look first at your Requirements (both internal and customer), but consider known failure modes and their failure rates (when known).

--> the processes shall be carried
--> out by qualified operators and/or shall require continuous

Key words in bold. So - you only have to have one method.

But again, as I said earlier, I would consider (as always) both internal and customer requirements (not to mention some common sense).

--> monitoring and control of process parameters to ensure that
--> the specified requirements are met.
-->
--> The requirements

What are your requirements - both internal and customer?

--> for any qualification of process operations,
--> including associated equipment and personnel (see 4.18),
--> shall be specified.

You addressed this in your training/certification program. Right? With regard to requirements, right?

Did you over do it? Probably a little bit.

[This message has been edited by Marc Smith (edited 08-27-98).]