Welding - Is welding always a special process per ISO9001:2000 Clause 7.5.2

[piney]

A special process is any process where the final results cannot be verified through a destructive test.

Welding will always be a considered a special process.

 

[Sidney Vianna]

Welding will always be a considered a special process.

WRONG.

Read my previous post on this thread and you will see why.

 

[AndyN]

Re: An example of a non-special weld process.

An architectural company orders some rectangular plant boxes made of stainless steel plates, duly welded. The only requirements are the dimensions, workmanship and the need for the plant boxes not leak around the seams. The first two requirements can be easily verified, so can the last one, by a leak test. A welding process was used, but all required characteristics can be easily and subsequently verified. No need to validate this weld process.

Since this thread has been revisited, I've had a chance to rethink this. I belive you have missed the point Sidney and use an example to justify your answer!!

It is true to say that the application is a low risk, but I still stand by that you cannot look at a weld and know that it's O.K. I remember taking a bike back to the store where the top tube joint had been brazed and everything 'looked' O.K except they'd changed the tube material and the tube was brittle! It cracked and broke after only a short time.

Had the process been treated properly, care would have been taken to ensure the tube was to the correct specification, before brazing! A step at the heart of a 'special process'.......

I'm going to have to disagree with you..........the criticality of product application only drives the type and amount of controls required.........

 

[Sidney Vianna]

Re: An example of a non-special weld process.

but I still stand by that you cannot look at a weld and know that it's O.K.

If you carry that line of thought, EVERY process requires validation. In the example I gave, the only requirements for the weldment are: aesthetically pleasing (workmanship) and leak-proof.

If you still think the welding I offered in my example requires validation, then I ask you: what kind of validation would you expect to see for that welding process, equipment and welder?

 

[AndyN]

Having made some welds, myself, which looked good (to me) but when subsequently tapped hard on a vice or steel table top, broke - I'd have to say 'Competencies' of the welder!!

Control of the material (we don't want the thing to fall apart if the metal becomes brittle!). You have extended it to the equipment, but that could fall under welder qualification.......couldn't it?

While I think you are over reacting to the premise, by saying all processes are this way - which I hope you aren't - my memories run to the Chrysler Neons running around which have paint stripping off them.......sure it looked good when it left the factory, was painted to the correct thickness, gloss etc., but the undercoat wasn't correct and didn't 'key' the top coat!!!

In fact, yes, maybe all processes should be validated! Isn't that the basics of (correctly) using tools such as SPC? Isn't the reason why Toyota has done so well with its manufacturing effectiveness and efficiency due to its relentless pursuit of process verification and validation, instead of QC?

 

[Sidney Vianna]

I am very surprised with your position, Andy. I thought that you, like me, was a firm believer in seeing ISO 9001 through the added-value prism. To make a bold statement such as all welding processes need to be validated is nonsensical, in my opinion.

For example, the welds performed to erect scrap-metal sculptures such as this one need no validation whatsoever.

.

Try as you might, you will never going to be able to convince me that this welding process needs to be validated.

 

[w_grunfeld]

Sydney,
I feel compelled to cast my vote to Andy on this...The very particular example that you bring to prove your point is correct but hardly relevant.
In an "artist's" shop where sculptures of this kind are created, indeed there is no need for weldment certification, or shall we say to a much lesser degree (the welder should still have to have some basic training but that is not the issue.) In a situation like you describe there is no need for a quality system at all, ISO 9001 or otherwise, so your example is not relevant.
My late mothers cooking wasn't ISO9001 certified and I CAN assure you it was a lot , tastier, fresher, healthier than that of any ISO9001 certified caterer's. So what does that prove? NOTHING.
An artist would object in principle to submit to any type of process control. Each piece is unique to be "art"...., and I presume he would take pride that no two pieces he creates are alike.
I think we are discussing here real situations, manufacturing processes (not art) and as such Andy is absolutely right. I worked once in a place that made electronics, we manufactured our own enclosure of aluminium and it had a welds. The welder was pretty experienced and the boxes looked good, none of it fall apart.
When a more knowlegdeable (aerospace) customer came along, our argument that it wasn't any structural integrity concern as it was not an aiframe part he didn't buy the argument, and insisted that if our welder was that good let him take the certification exam. Guess what? He did and he failed....
In an industrial environment you can either have a certified process (in which case you can't limit the scope to sculptures or flower pots only) or you don't .
Best regards,

 

[Paul Simpson]

Re: Welding - Is welding always a special process per AS9001 Clause 7.5.2

Sydney,
I feel compelled to cast my vote to Andy on this...

... and I have to support Sidney. Both schools of thought know what they want from this. The thread developed with the following post by Andy:

Welding has always been considered a 'special process'. The quality of a weldment, in its entirety, cannot be determined through any amount of fancy inspections.

Welding requires many controls to be in place; operator/welder certification/qualification, welding materials controls/storage/conditioning, weldment material preparation/preconditioning, approval of welding procedures, and so on.

Without these kinds of controls, it is highly likely that latent defects, un-inspectable by the techniques mentioned, can be introduced.

Sidney's example took this (perhaps to the extreme) by showing how welding can be part of a 'manufacturing' process but no requirement for 'special' controls.

The very particular example that you bring to prove your point is correct but hardly relevant.
In an "artist's" shop where sculptures of this kind are created, indeed there is no need for weldment certification, or shall we say to a much lesser degree (the welder should still have to have some basic training but that is not the issue.) In a situation like you describe there is no need for a quality system at all, so your example is not relevant.

The reason there is no need for 'special' process controls is solely that welds do not have any significant engineering function (beyond resisting gravity ).

An artist would object to try to submit to any type of process control. Each piece is unique to be "art"...., and I presume he would take pride that no two pieces he creates are alike.

I'm also not convinced with the point on process control / all pieces being one offs. Most artist have good control of their processes - there is the story of the pope selecting architects based on their ability to draw a perfect circle here. All pieces may be commissioned as one off but that doesn't mean the artist wants the quality of the weld to vary.

I think we are discussing here real situations, manufacturing processes (not art) and as such Andy is absolutely right. I worked once in a place that made electronics, we manufactured our own enclosure of aluminium and it had a weldds. The welder was pretty experienced and the boxes looked good, none of it fall apart.
When a more knowlegdeable (aerospace) customer came along, our argument that it wasn't any structural integrity concern as it was not an aiframe part he didn't buy, it and insisted that if our welder was that good let him take the certification exam. Guess what? He did and he failed....
In an industrial environment you can either have a certified process (in which case you can't limit the scope to sculptures or flower pots only) or you don't
Best regards,

The question in the case of your customer is what engineering function he wanted the box to have - if there was none then there is no need for qualified welders, certificated welds, controlled processes. The weld just needs to look OK.

Neither Sidney or I is arguing that engineering welds don't require 'special' controls - only that not all welds are engineering welds. Those that are cosmetic can be controlled visually.

 

[AndyN]

I had started a response earlier, but got bumped off line.......

I'm not trying to convince anyone! I'm merely taking a different point of view, Sidney!

My experience, training (as an apprentice tool maker etc.) and education as a mechanical engineer tells me that many products from 'special processes' cannot be assured by visual controls. Indeed, I believe that we're 'over blowing' the validation steps - to make them seem ridiculous and, hence, unnecessary!

Let me put it this way. Not everyone can weld - so it's important to know someone's competent - that's part of validation, isn't it? You can't weld all types/grades of materials together successfully. Control of materials, is part of validation isn't it? Similarly with equipment - if the (we'll say electrically powered) welder is underpowered, or incorrectly set, that's important to welding isn't it? So, in general terms, all these things have to come together - and often proven by a test or sample weld, before production begins. Isn't that part of validation? It has to check all these criteria though, not just that the weld looks good!

Similarly, I've (easily) pulled apart wire crimped terminals which looked fine. Luckily, I found them, since they were a electrical safety barrier box application. The people who had made them and inspected them were not qualified to know any better!

To Willy's point, I've seen some good looking food, which subsequently turned out to be nasty tasting!

 

[Sidney Vianna]

Indeed, I believe that we're 'over blowing' the validation steps - to make them seem ridiculous and, hence, unnecessary!

And my contention is: anytime someone makes generalized, all-encompassing, dogmatic statements, such as ALL WELDING PROCESSES NEED TO BE VALIDATED, they are bound to be wrong.

I think you are mixing the skill to be a welder and validating the welder. In my previous example, if the artist who builds such scrap-metal scupltures was forced to validate his welding process, it would be a crass example of non-value added bureaucracy that many auditors and consultants would demand, because they can not think out of their boxes and heard once that ALL WELDING PROCESSES NEED TO BE VALIDATED.

Remember, common sense still applies.