TS 16949 - 7.5.2 - Process Validation - Criteria for review and approval of processes

[Morpheus]

Process validation 7.5.2

I'd like to know how third party auditors are auditing this clause requirements. For example, TS requires to define criteria for review and approval of processes. But could this criteria be as strict or loose as you want ?
The approval of equipment and qualification of personnel is also left to the user taste.. or do auditors are requiring specific benchmarks or targets ?
At our company we developed our own set of rules on this, and we'd like to know how other companies are handling this.

Regards
Morpheus

 

[howste]

Yes, you decide your own criteria. Who would know better how to validate your processes than you? I believe it's impossible for auditors to have set criteria that they look for in every company. They should evaluate the effectiveness of what you have established.

 

[Morpheus]

Thank you, howste. What we have done is to develop a checklist with apqp topics and some other relevant stuff. We have this as a primary cover and attached support info.

Regards

Yes, you decide your own criteria. Who would know better how to validate your processes than you? I believe it's impossible for auditors to have set criteria that they look for in every company. They should evaluate the effectiveness of what you have established.

 

[howste]

Sounds good. By the way, much of the required process validation information may already be included in your PPAPs and/or Run at Rate (if you supply to GM).

 

[p_tww]

Sounds good. By the way, much of the required process validation information may already be included in your PPAPs and/or Run at Rate (if you supply to GM).

As TS2 requirement, all process should be re-validated. whether the frequency could be defined by company self? Whether process parameter should be re-validated at their upper/lower specification? Whether the process could be seperated as several bigger process, not detailed small process?
I am not English spoker. Hope this could be understood. thanks.

 

[howste]

1) You can decide your own revalidation frequency unless your customers require something different.
2) I believe that when possible you must validate at upper/lower limits. If all you do is test nominal conditions, you don't know how far you can deviate from those conditions without problems. Testing at the limits should ensure that anything in between will also be good.
3) I'm not sure I understand your last question. Could you give an example of what you mean by "Whether the process could be seperated as several bigger process, not detailed small process?"

 

[p_tww]

1) You can decide your own revalidation frequency unless your customers require something different.
2) I believe that when possible you must validate at upper/lower limits. If all you do is test nominal conditions, you don't know how far you can deviate from those conditions without problems. Testing at the limits should ensure that anything in between will also be good.
3) I'm not sure I understand your last question. Could you give an example of what you mean by "Whether the process could be seperated as several bigger process, not detailed small process?"

Howste, thanks. you are right with what I am thinking.
for the 3rd questions. example: one company have many processes for one kind of product , such as machining, welding,grinding,assembly etc. you know for assembly process, it include many steps. Should company validate each step or just validate whole assembly process?
Hope this clear.

 

[howste]

If you can validate the entire process without validating each step, there shouldn't be a problem. But - it's probably difficult to do without testing variables at each step.

A company I used to work for made suspension struts for automotive. For the piston rod there was heat treating, straightening, machining, two grinding operations, chrome plating, and final buffing. If the product came out good, then the process was good. But, we had to find out what straightening parameters would allow us to get good product after grinding, what grinding parameters allowed us to get good product after plating/buffing, etc.

One thing that you may consider is that machine capability can be determined in a way that is not product-specific in many cases. I believe that would be a way to reduce the amount of work validating processes and should be sufficient to meet the requirements.

 

[p_tww]

Thanks, Howste. totally agree.