A number for PQR

Date 05-31-2012 11:55

I always thought the A-No should just go away. No matter what you do it will be wrong. If you test a representative groove you will never get the same number throughout due to dilution, as you say. If you test some other configuration like a chem pad it will not even be close to representative.
To me it just adds needless confusion.

By 803056

Date 05-31-2012 16:04

I think the A-number is just one way to keep idiots from attempting to weld various types of steels using different filler metals base on the assumption that if the F-number and P-number is not affected, all is well.

Consider a WPS qualified austenitic stainless steel welded with ER308 filler metal using the GTAW process. If A-number is disregarded, an idiot would conclude the base metal could be welded with ER70S-2 because the P-number isn't changed, the F-number has not changed, and A-number is no longer a factor. Oops! Foiled again!

A-number can be very misleading unless the chemical analysis is performed on a recast nugget of the weld cross section to ensure the chemistry is homogenous. If the multiple layer weld is sectioned and chemistry is taken from the deposit at the edge of the groove, it will be quiet different than the chemistry determined for a layer extracted from the top third near the centerline of the weld.

I guess that is why good welding consultants are in demand.

Best regards - Al

By js55

Date 06-01-2012 13:18

Al,
Agreed. The A-No. system cannot be eliminated without some adjustment to the F-No system. They are intimately related. But the F-No. system is ued for control of non ferrous materials With some adjusmtent it could be used to control all materials.The odd thing is people are concerned about controlling what goes in to a weld, i.e. F-No. So therefore the F-No should be the control mechanism. As least this is the argument I am making. If we can do it with non ferrous we can do it with ferrous with some improvement.

By 803056

Date 06-01-2012 13:38

That is true to a limited extent.

The issue is when dissimilar metals are joined and when different filler metal compositions come into play. An example of what I have encountered would be when working with the aluminum alloys. For example: welding 6061-T6, everyone's favorite alloy. The engineer can use several different filler metals, each with specific mechanical and chemical properties. The mechanical properties govern, i.e., the weld usually out performs the base metal, but that isn't always the case when joining different aluminum alloys to the 6061-T6.

One welding standard I used to use when welding aluminum alloys required a chemical analysis of the weld. Changing groove details changed the chemistry and it also affected the properties of the weld, i.e., the strength of the joint. It was a dilution issue. Any changes in the filler metal selection or groove detail required a new PQR to demonstrate the combination provided the properties required. It made sense then and it still makes sense to me now. I find it interesting that the chemistry issue has fallen by the wayside. New people, new opinions. Not that I find the new opinions are always correct. This is one case where the chemistry made a difference in the ability of the joint to function as intended. What works with a V-groove does not necessarily work with a single pass square groove.

I believe one of my good friends says it best when he refers to Aluminum as "almost a metal".

I blame it all on the ASME influence. :eek:

Best regards - Al

By js55

Date 06-01-2012 16:48

Al,
I think one of the reasons that weld deposit chemistry has fallen by the wayside, and actually the primary reason for me, is that testing does not tell us what the chemistry actually is. Its a poor approximation whether chem pads or grooves. The weld deposit is much more heterogeneous than the A-No system assumes. And dissimilars don't make it better, they make it worse. If you do a dissimialr between carbon steel and P22 using B3 filler, I can garantee you are going to have deposit chemistries anywhere from A-1 to A-4(and the A-1, A-2, A-3 chemistries were mechanically tested just as sure as the claimed A-4). So just exactly what has a concern for deposit chemistry achieved other than convincing us of something that doesn't exist.
An interesting if little known result of this is that if you find you have violated your procedure by PMI just test in another place you'll find it eventually. But if you control the filler going in, test it, its good, now you have something real you can wrap your mind around and engineers can apply.
Now, when the A-No system was imposed I believe they understood this quite well but figured it an adequate approximation. Maybe so back then, but I think controlling whats going in is far more robust than guessing and approximating the result.
This is not unlike the old concern for ferrite in weld deposits. If you miss your ferrite you can increase or decrease your cooling rate and viola' compliance. Which is exactly what filler metal companies did for years and years. They knew exactly (well, maybe not exactly, the diagram guys have fought this for yesr) what cooling rate would match the chemistries they ended up with in their heats. It was a game.

By 803056

Date 06-01-2012 23:49

Once again we are almost in agreement. Simply taking chemistry from a random location doesn't represent the average chemistry. Simply knowing what filler metal, i.e., controlling the filler metal doesn't determine the chemistry of the final weld deposit because it doesn't take into consideration dilution. The chemistry along the edges of the groove will be more highly diluted than the center of the multiple pass weld. The only time there will be homogeneity is with the single pass weld. At best the determination is an approximation even when using a remelted plug for the chemistry. Whether it should be an essential variable is debatable. However, I believe I understand the intent and rational for the A-number as a consideration. Why it isn’t applicable to other metals, i.e., nonferrous metals, confounds me. So many nonferrous metals are joined with dissimilar metals where the amount of dilution is essential to success or failure. Yet, the A-number isn’t an essential variable when meeting the requirements of Section IX when joining nonferrous metals.

This is where and when the welding consultant earns his fee. Whether you are using a carbon equivalency equation, the WRC for approximating Ferrite Numbers, or chemistry to determine the A-number; each is simply a tool to help the engineer anticipate or to devise a solution to a potential problem. The information provided by the tool gives the engineer a warm fuzzy feeling that all is well in Welding Land or gives the engineer a cold shiver as a warning there is a potential problem. Part of the fun is learning which tool to use and what it is trying to tell us. The results are rarely black and white, but shades of gray.

Best regards -Al

By js55

Date 06-04-2012 14:21

Al,
We are close in agreement but I would make one more point. I do not see the A-No as a tool to help the engineer to anticipate or devise a solution in that the A-No is for qualification purposes only. It is not intended to be applied after qualification or for any service assessment or anticipation. It cannot be used for that (and would not be suggested to be used for that by Section IX) since that is not its purpose. Service engineering decisions will be made WITHIN the A-No. You wouldn't make engineering decisions based upon P-No. You would make it upon precise alloy. A-No is simply a qualification grouping as is P-No. A-No essentially gives you no engineering information at all.
Also, any problems to be solved in service can be just as easily done by what goes in to the weld, the filler, as opposed to a guess on the result. Engineers make filler decisions everyday. If a more precise measure of weld metal chemistry is needed then the A-No concept is no help. The A-No by definition is essentially nominal and average. I have never made an engineering decision based upon A-No but I have made hundreds of engineering decisions based upon filler.

By 803056

Date 06-05-2012 17:12

As usual you have taken a position that is hard to argue.

The A-number is for the purpose of qualifying the WPS and I cannot argue that point. I guess what I was trying to point out was that the A-number alone isn't going to provide assurance there will not be a problem. Dilution must be considered when it comes time to analyze the application especially when working with dissimilar metals.

I like your argument because you are more precise and clear with your explanation. Nice job.

Best regards – Al

By js55

Date 06-06-2012 11:58

Al,
Compliments from yourself are always appreciated.
It is seldom the case that we are far apart and we spend most of our time debating details. In this case my position is a developing one and it may be the case that the A-No system is going to come under some scrutiny in the near future. I favor eliminating it and utlizing an F-No. with suffix system similar to the P-No system, for example P-15E, and to be applied to non ferrous as well as ferrous. It stands to be met with some very strong and cogent opposition. The A-No system has been around for some time, it works pretty well, and has tremendous inertia. Emphasis on pretty well. Its going to be interesting to watch this evolve.
Section IX will be undegoing a lot of change in the near future. We may add this to the list.

By welderbrent (*****)

Date 06-06-2012 13:19

And I always feel like a silent benefactor of these little discussions while you two hash out the details and clarify your positions. Lots of good info and application here.

Have a Great Day, Brent

By js55

Date 06-06-2012 20:15

Brent,
The idea of other benefactors is always present to mind. I certainly don't attempt to take on Al's superior knowledge because I'm a masochist. :lol:

By welderbrent (*****)

Date 06-06-2012 20:33

Just for the record, even some of the more argumentative posts in other threads are still educational to those of us who have so much to learn. It becomes a matter of sorting through the info as presented, getting into the code books and other references for oneself, and coming to, what is prayerfully, a proper code application interpretation.

There are so many issues within the assorted codes that we have a great diversity of opinions about. Bottom line is to sort some of it out with the Engineer of record for the job, the Building Official on the job, the Inspector or Record on the job, the onsite inspectors, the fabrication shop, and the erectors. Some times all of them at once.

But I sure appreciate all the time and vast amounts of knowledge shared here that at least gets us started down the right path. There is great wisdom, and safety, in a multitude of counsellors. Even if they are all wrong. Because it makes us think for ourselves and go to the powers that be that are responsible for the outcome.

Then, there are those times when we at least try to go to the Code Committee's and find out what they intended when the code was written. But the Engineer can still apply his own interpretation.

Have a Great Day, Brent

By 803056

Date 06-07-2012 16:22

The Forum is akin to the courses I teach, no one leave knowing less than they knew upon arrival.

That applies equaly to me as well as other people. You just need to keep an open mind.

Best regards - Al

By js55

Date 06-07-2012 19:34

Al,
No but you may leave 'feeling' you know less because of a realization that things are much bigger and more complicated than you previously knew. :grin:

That happens to me all of the time.

By 803056

Date 06-08-2012 20:04

That, in its self, is an education.

Al