Is there a "list" of what are termed Special Processes that have to be validated?

[Paul Simpson]

There is a list in the guidance from GHTF (GHTF/SG3/N99-10:2004 edition2).

Please escuse my ignorance but:

• Who are the GHTF?
• What is this document?

 

[Jim Wynne]

Please escuse my ignorance but:

• Who are the GHTF?
• What is this document?

Global Harmonization Task Force

 

[Paul Simpson]

I was impressed when I saw 'Global Harmonization Task Force' :mg:

But then I saw it was just another quasi standards body with another range of opinions!

 

[Helmut Jilling]

Re: Agree with the principle

It all comes down to risk. The reason for ISO requiring process validation for "special processes" is that you cannot demonstrate they conform any other way (within reason). I agree there are many reasons why you might want to validate a process even though there is no requirement. Process repeatability and the ease of identifying failures are factors you might consider.

In many industries there is little process planning and even less process verification / validation. Not an ideal situation but that is the fact.

Excellent points, Paul. The reason ISO does not give a definitive list of what are specialprocesses, is it varies. One example given in ths thread was plating thickness. For many, that is a special process because they can't validate it by normal measuring methods. But, if a company buys a meter to measure it, then it is not a special process for them.

So, the ISO definition is any process you cannot measure or validate but ordinary or other means.

Maybe it would be simpler if we say it this way, "if you can measure it, it is NOT a special process in your system. If you can't inspect it, then it probably is."

 

[somashekar]

Re: Agree with the principle

Excellent points, Paul. The reason ISO does not give a definitive list of what are specialprocesses, is it varies. One example given in ths thread was plating thickness. For many, that is a special process because they can't validate it by normal measuring methods. But, if a company buys a meter to measure it, then it is not a special process for them.

So, the ISO definition is any process you cannot measure or validate but ordinary or other means.

Maybe it would be simpler if we say it this way, "if you can measure it, it is NOT a special process in your system. If you can't inspect it, then it probably is."

Going further on this, if I have a hardness tester to check the hardness of the cured surface of a 2 part epoxy potting, and the measured number with it meets my specification as well as the cured epoxy nominal specifications, then it is not a special process .... ?
I beg to disgaree. Plating thickness or hardness is just one parameter of the process. How will we verify the plating adhesion to substrate or the epoxy adhesion to the substrate? Defects from such processes may become evident over time and just a single parameter measure may not be sufficient to say that the process was optimum. A good scientific thought for a process will decide if it is special and such processes indeed needs a validation, revalidation when any change occure in the process inputs. Otherwise you welcome complaints (mostly customer complaints) which are much harder to crack and end up with validation finally as the corrective action.

 

[Helmut Jilling]

Re: Agree with the principle

Going further on this, if I have a hardness tester to check the hardness of the cured surface of a 2 part epoxy potting, and the measured number with it meets my specification as well as the cured epoxy nominal specifications, then it is not a special process .... ?
I beg to disgaree. Plating thickness or hardness is just one parameter of the process. How will we verify the plating adhesion to substrate or the epoxy adhesion to the substrate? Defects from such processes may become evident over time and just a single parameter measure may not be sufficient to say that the process was optimum. A good scientific thought for a process will decide if it is special and such processes indeed needs a validation, revalidation when any change occure in the process inputs. Otherwise you welcome complaints (mostly customer complaints) which are much harder to crack and end up with validation finally as the corrective action.

Your comments reinforce why ISO does not provide a list of "special processes." It varies from company to company, depending on their needs and what equipment they use. You have gone into specific details on one situation.My example was meant to be a simple example - how a process could be "special" at one company, and not special at another.

 

[achorste]

Just an additional point on validation - just because you can measure it doesn't mean it's not a 'special process'. Hardness for example - you can easily check the hardness of a material at any point - however that value only applies for that one point. You can measure it, but how do you know the entire product is at this hardness without either validating the process or inspecting every point on the surface?

 

[Stijloor]

It can also depend on the industry. Automotive (ISO/TS) requires all manufacturing processes to be validated.

Stijloor.

 

[Helmut Jilling]

Just an additional point on validation - just because you can measure it doesn't mean it's not a 'special process'. Hardness for example - you can easily check the hardness of a material at any point - however that value only applies for that one point. You can measure it, but how do you know the entire product is at this hardness without either validating the process or inspecting every point on the surface?

I think if you apply this logic too stringently, you will end up making most processes"special." I don't think that was the intent ISO had.

Of course, processes need to be validated when developing them. This section of the standard is simply saying, if you cannot validate the product by inspection, then you have to control and validate by inspecting and monitoring the parameters.

Beyond that simplistic statement are many details which must be considered by companies when setting up their systems.

 

[achorste]

Beyond that simplistic statement are many details which must be considered by companies when setting up their systems.

True - consideration has to be given to risk, benefit and cost implications.