ISO 9001:2000 Clause 7.5.2 Validation of Processes - Level II Response

[apestate]

I'm not sure how to read this requirement, at all. I struggle with the following sentence under:
7.5.2 Validation of processes for production and service provision

"The organization shall validate any processes for production and service provision where the resulting output cannot be verified by subsequent monitoring or measurement."

If I make widgets with Widgetmakers and control my production and service provision in 7.5.1, I wonder what output cannot be verified by subsequent monitoring and measurement.

my imagination becomes too active when I read a. b. c. d. e. requirements, because I cannot think of any processes where deficiencies should become apparent only after the product is delivered, or any output of production & service that cannot be verified by monitoring or measurement.

I think of 7.5 as my little factory, and it is a central point in the midst of the long string of the ISO 9001 requirements. everything feeds into it, hahaha. like a cartoon for the auditor, a fake process map.

can someone please explain 7.5.2 briefly in terms of application to a small hardware manufacturer on quote contracts? I can think of nothing!

Also, is it wrong to type out sections of the standard on this board when I'm asking questions like this?

 

[CarolX]

"The organization shall validate any processes for production and service provision where the resulting output cannot be verified by subsequent monitoring or measurement."

can someone please explain 7.5.2 briefly in terms of application to a small hardware manufacturer on quote contracts? I can think of nothing!

Is your hardware heat treated, is it plated? How do you verify this features? Certain process can only be controlled and verified by strict monitoring of the process.

Hope this helps clear it up a bit.

CarolX

 

[Cari Spears]

Here's an example - I used to work at a place that bonded friction material to steel cores and bands. (transmission and brake parts). The way they verified the bond was to bend the pieces over a mandrel and then try to peel off the friction material. We did destructive bond testing 2X per shift, and in the interim we monitored process parameters that had been established during R&D of the process. Oven or platten temperature, pressure, cycle time, and viscosity of the adhesive were some of the process parameters monitored for product that could only be verified by destructive testing.

 

[apestate]

Carol:

None of our product realization practices can ONLY be verified by close monitoring of the process.

All heat treating, plating, oxidation or other such process processing is done by approved suppliers. They provide certificates of conformance. If I start to re-check their work, will I stop before I do the metallurgy on the steel coming in from the mills?

The only thing I can reasonably say is not verified by subsequent monitoring or measurement is shipping. 7.5 goes through release & delivery activities, so I could have JUST ONE THING for 7.5.2. Ship a box of parts to Chicago and back to see if they get dinged up?

Is 7.5.2 ever a possible exclusion, and in what sort of manufacturing would it be excludeable.. excludable.. ex.. to leave it out?

 

[Wes Bucey]

in-process validation versus after-the-fact

Carol:

None of our product realization practices can ONLY be verified by close monitoring of the process.

All heat treating, plating, oxidation or other such process processing is done by approved suppliers. They provide certificates of conformance. If I start to re-check their work, will I stop before I do the metallurgy on the steel coming in from the mills?

The only thing I can reasonably say is not verified by subsequent monitoring or measurement is shipping. 7.5 goes through release & delivery activities, so I could have JUST ONE THING for 7.5.2. Ship a box of parts to Chicago and back to see if they get dinged up?

Is 7.5.2 ever a possible exclusion, and in what sort of manufacturing would it be excludeable.. excludable.. ex.. to leave it out?

As Carol says, this is not mysterious, just plain old everyday stuff.
Imagine your process produces a sealed "black box" with lots of little 'things' inside. Once the box is sealed (the output), you can NOT validate the number of 'things' or whether they are correctly fastened/connected. Therefore, you validate various aspects while the black box is in process, BEFORE it is sealed.

Given that kind of scenario, I'm sure you can imagine many other equivalents to a "black box."

 

[howste]

I don't agree with the "black box" approach in many cases. If I verify that the number of "things" is correct and properly fastened/connected before I seal the box, then it still doesn't fall under 7.5.2.

The processes that really need to be validated are ones that you can't verify the outcome after the process is finished (not the entire assembly). Without destructive testing, can you verify the strength of a resistance weld after it is completed? Heat treating, plating, etc. that Atetsade mentioned above would probably be candidates if they were done in-house. Of course, don't forget to include controls for the outsourced processes.

My experience has been that it's possible to exclude 7.5.2 in many companies with various products. Most I've seen don't exclude it though, preferring to include a contingency plan in case a new process is developed that requires it.

 

[Randy Stewart]

Without destructive testing, can you verify the strength of a resistance weld after it is completed?

Hit the nail on the head with that one.
As for the "black box", if it works you validated the process. With welding one way you can do this is by the welding heirachy.
Weld coupons (most frequent - tip changes, pressure changes, etc.) validate pull tests frequency
Pull Tests (medium frequency - x amount of parts, etc.) validate tear down frequency
Tear Downs (least frequent - set-up and end of run) verifies the overall process

 

[CarolX]

All heat treating, plating, oxidation or other such process processing is done by approved suppliers. They provide certificates of conformance. If I start to re-check their work, will I stop before I do the metallurgy on the steel coming in from the mills?

All the spec says is you have to validate in some manner. You utilize approved suppliers for your special process. You have some method for approval of these suppliers, and the provide you with a CoC. I would say you have covered the requirements.

JMHO,
CarolX

 

[Rachel]

Okay, so on a similar line...

Hi everyone...
So I'm struggling with this one too. Ideally, in a perfect world, everything would fall to process validation over production control...for the sake of catching problems at the earliest possible time, minimizing rework, so on, so on, blahblahblah. I don't know what I'm getting at here...

I guess the question is this: is a destructive test still considered product testing, or is that process testing? I mean, you obviously can't destructively test *all* pieces...but then, you don't necessarily test *all* pieces anyway. I'm just having a hard time thinking of an example where there is no way (destructive or otherwise) to verify the output of a process.

Can anyone shed some light? It's pretty dark in my world.

Cheers all,
-R.

 

[Randy Stewart]

You can do it

Rachel,
You need to define what the testing is trying to do. Here's an example with welding:

The tear-down tests are used to validate the process.
The pull tests are used to validate the tear-downs and on-going performance,
the weld coupons are then used to validate the setups.

Another way would be:

Weld coupons show you can set the process up properly,
the pull test show that you can operate within process parameters,
and the tear downs show that you have a compliant process.

Does that help?