Not logged inAmerican Welding Society Forum
Forum AWS Website Help Search Login
Up Topic American Welding Society Services / Certifications / Aluminum 6160-T6
- - By TimGary (****) Date 06-17-2013 17:52
I'w working on the qualification of some T6 Alum. procedures.
I'm trying to do CJP on 3/8" plate.
I've tried with and without bevels, pulse and spray, hot and not so hot.
My samples all look great, but none survive bend testing, with most breaking at the toe of the weld.
I'm using 4043 x .047 wire and 100% Argon.
21-25 Volts,  200 - 450 IPM WFS, 10-15 IPM Travel Speed.
Working to AWS B2.1.

I was warned by the guy who does my tensile tests that 80% of the tensiles he's pulled have failed, even though B2.1 only requires a minimum of 24 KSI tensile.

Is anyone familiar with this willing to give me some advice to help speed up my R&D?

Thanks,
Tim Gary
Parent - By bozaktwo1 (***) Date 06-17-2013 20:17
You're going to have trouble getting those tensiles with 4043.  4.7.5 (2) allows the use of the minimum tensile strength of the weld metal.  If your brand of 4043 pulls less than your 6061 "as-welded," I would go that route.  As for your break at the toe, it's very important to get the coupons down to the correct thickness, and to make sure that the surfaces are flat and smooth.  I don't recommend the plunger-type bending jigs.  Use a rotary bender; it distributes the force of bending more evenly.  A problem I found when doing these tests is that the HAZ is annealed, probably near to O condition, so it's more likely to bend sharply at the toe of the weld.  It's less likely to do this if clamped into a tight rotary jig.  You may want to try side bends in lieu of face and root bends.  Finally, I'll say what doesn't need saying: clean, clean, clean...and when you think it's clean, clean it again! :grin:  Good luck, and let us know how it comes out.
- - By 803056 (*****) Date 06-17-2013 22:16 Edited 06-18-2013 00:52
Oxides as a result of insufficient cleaning or failure to limit the time at temperature are typically the culprits leading to low tensile strength.

Degrease the parts using virgin acetone followed by an alcohol flush. Then wire brush the components using a stiff stainless steel wire brush. Brush the parts using strokes in one direction as you would if you were cutting with a hacksaw. Just before welding, flush the parts with 90% isopropyl alcohol and then do not handle the parts with you bear hands. Always wear clean fabric gloves when handling the cleaned parts. Weld within the hour.

You need to be mindful of the interpass temperature. Use a fan to cool the parts between weld beads. DO NOT use compressed shop air to hasten the cooling rate. DO NOT preheat the aluminum before welding. 6061 is a heat treatable alloy that will over age if held at temperature for an excessive amount of time.

You should be able to pass the required tensile tests using either 4043 or 5356 filler metal.

Aluminum: almost a metal.

Good luck - Al
Parent - - By TimGary (****) Date 06-18-2013 11:28
Thanks for the replies guys!

Any suggestions as to heat input?
I'm thinking medium-high, about 24 V and 350 IPM wire...
Pulse or Spray?

Tim
Parent - By 803056 (*****) Date 06-18-2013 12:02
Spray mode only.

Al
Parent - By bozaktwo1 (***) Date 06-18-2013 16:58
The beauty with the "almost a metal" is that it can be welded in spray in all positions.  Almost a metal.  Diggity. :grin:
Parent - - By tom cooper (**) Date 07-01-2013 16:51
Hey Tim,
I don't often get to post recommendations as I seem to be forever in need of help myself, but I know a lttle about 6061.
You should easily achieve 24ksi tensiles with 4043 as long as you don't char the he$$ out of it. Even my most porosity ridden sponge like welds will exceed 24ksi so don't worry about that.
For bend testing, be sure you (or your lab) anneals the coupons properly and the bend mandrel diameter is correct, you will have NO problem passing these mechanicals.
Be grateful you don't have to pass x-ray.

As for heat input, this is not an alloy that is advertised with any heat input limit such as A514 or HY steels so you won't find that published anywhere on this earth, but I calculate this for all my PQR's anyway and my numbers run consistently on the order of 30kJ/inch maximum (thats a tig number, but a number for you to use as a baseline).  

Hope this is of some help and sorry if I was late to the discussion.
Parent - - By 803056 (*****) Date 07-02-2013 15:11
Are qualifying the WPS with or without PWHT?

Al
Parent - - By tom cooper (**) Date 07-03-2013 20:30
Hi Al-
Was your question directed to me ? or a convenient place to attach continued dialogue with Tim?

Regardless, my answer to Tim was with regard to the 'as-welded' condition since the context of his question was with the use of 4043. otherwise pwht (depending on intent) wouldn't be very effective unless 4643 was used.

But since you brought it up, why did you bring it up?
Parent - - By 803056 (*****) Date 07-04-2013 00:28
Hi Tom;

My comment was directed to you. You made the statement that bend samples needed to be properly annealed before bending. That would be considered a PWHT and it would be an essential variable. As such, the specific PWHT would have to be listed on the PQR.

I was and still am interested in the PWHT you used or recommend. 6061 is time at temperature sensitive and "easily" overaged. Overaging causes the alloying constituents to precipitate to the grain boundaries and doesn't do any favors for the tensile strength. So, I was interested in what and how you "anneal" your samples before bending them.

You can see why I am interested. There is always something to learn here in the Forum.

Best regards - Al

Happy 4th of July!
Parent - - By tom cooper (**) Date 07-04-2013 10:45
Good morning Al and happy 4th back to you.

My annealing comment is based on Tim's governing qualification document, B2.1 and of course only applies to the bend test specimens (I didn't say the tensile coupon should be annealed).

Figure II-5A, -5B and -5C tells us to 'anneal' M23 bend specimens when 3/8 thick are used with a 2 1/2" mandrel or plunger diameter. This specification doesn't seem to instruct us on how to anneal the specimen and that opens up opportunity to anneal to our most ductile advantage. But to avoid the appearance and possible accusation of cheating, I would add a footnote to the PQR and state that bend specimen was annealed per guidance for M23 base metals in AWS D1.2 Clause 3.8.1.7.

There is good reason for AWS to allow the anneal practice (for bend tests only) and that is because the elongation of the 4043 weld material is >>> greater than the HAZ or the unaffected base material of M23 alloys, so most of the stretch will occur in the weld filler, i.e., there isn't equal sharing of strain between the base and filler and so we will likely split at the HAZ which behaves much like a pivot point where the ductility is least. 

Because the specification allows the anneal to be introduced as p/o the bend testing regimen, it is not considered an essential variable in this circumstance.

Where above I said porosity riddled welds will still pass 24ksi tensile, if you are plagued with porosity, your bends may open up and stretch a discontinuity that exceeds allowable size (1/8").

This is my story and I'm stickin' to it:evil:, although I'd be very pleased to have your commentary and counter argument as I have never had the privilege of being challenged on it.
Parent - - By 803056 (*****) Date 07-04-2013 16:24 Edited 07-04-2013 16:34
Hello Tom;

I am confused. I have AWS B2.1/B2.1M:2009 in front of me. I cannot find the figures you are referencing, i.e. Figure II-5A, -5B, and 5C.

Using AWS B2.1/2.1M:2009 as reference,  Annex B, clause B2.4 does say that M-23 and M-35 (with exceptions for C95200 n C95400) must be no more than 1/8 inch thick when performing the guided bend tests. However, I see nothing in the text permitting the guided bend samples to be annealed.

Since I did not find the figures you referenced in AWS B2.1/2.1M, I searched further. I dug out my copy of AWS B2.1:2005. There I found Annex II clause II 2.4. It essentially has the same text as the text found in AWS B2.1/B2.1M:2009. Again I see no mention of annealing the guided bend samples in the text.

I search further and found the figures you listed. Figure II-5A; there is no mention of annealing the guided bend samples before bending in the footnotes, however, in the body of the table there are two listings for M-23. One listing is for M-23 in the as welded condition and a second listing for M-23 in the annealed condition. The bend radius is different depending on whether the coupon being tested is in the as welded condition or in the annealed (PWHT) condition. As expected, the property of elongation is better in the PWHT condition than it is in the as welded condition.  The improvement in the property of elongation is dramatic, 17.6% in the annealed condition versus 6.5% in the as welded condition.

I see the same information in AWS B2.1:2005 for Figure II-5B and Figure II-5C.

Continuing on in AWS B2.1:2005, I see that any aluminum alloy welded with F-23 filler metal is to be bent to the same radius as the M-23 alloys in the as welded condition.

My opinion is that the tables included in the figures of B2.1:2005 and AWS B2.1/2.1M:2009 are simply taking into consideration whether the PQR coupon is in the as welded condition or the PWHT condition (annealed) and use the bend radii that is consistent with the property of elongation for each condition. I do not believe the laboratory has the option of performing the guided bend tests in either the as welded or the annealed condition.

As I stated I my previous response, any PWHT is a change in an essential variable and would require a new PQR. In my opinion if the WPS is welded and tested in the as welded condition the appropriate bend radius for the 1/8 inch thick guided bend test sample is 2 1/16 inch. If the WPS is qualified in the PWHT condition, i.e., annealed, the proper bend radius is 2 ½ inches for the 3/8 inch thick guided bend test sample.  If the qualification test coupon is subjected to PWHT, i.e., the WPS is qualified with PWHT, all the tests must be performed in the PWHT condition and recorded as such on the PQR. I the WPS is qualified without PWHT, i.e., in the as welded condition, all the tests must be performed in the as welded condition.

By the way, just exactly how are you "annealing" the samples before performing the guided bend tests and are you reducing the thickness of the samples to 1/8 inch before performing the guided bend tests or are you bending them wth the full coupon thickness, i.e., greater than 1/8 inch?

That’s my story and I’m sticking to it! :confused:

Happy Fourth of July! Let the fireworks begin! :wink::smile:

Best regards - Al
Parent - By tom cooper (**) Date 07-05-2013 00:26
I think your notion that an engineered application which called for welding 6061-T6 and then PWHT annealing that weldment is a bit far fetched. No designer would ever (knowingly) specify T6 and then extract all of its strength by annealing the weldment. It would be a rare circumstance indeed and the AWS writers would not need to address that rare circumstance with a special mention of annealed T6.

For your argument to hold water, it would also have to apply to AWS D1.2 bend test prescriptions. Is it your opinion that para 3.8.1.7 does NOT allow me to anneal the specimen for bend testing purposes?   Before you answer that please note the introductory statement, "Welds on M23 base metals may be bend tested under either of two conditions, as-welded or annealed."

I might agree that B2.1 could have been a bit more explanatory. And I might agree that I am offering a generous reading of their placement of the word "annealed" immediately adjacent to M23 in the Figures II-5A, B and C but it is a reasonable reading based on trusting that AWS intends consistency between methods of D1.2 and B2.1. As proof of this please note that the bend testing diameter listed in B2.1 for a 3/8 thick specimen is 2 1/2" and the bend testing diameter listed for D1.2 annealed coupon is 6-2/3 x t which for a 3/8 thick specimen is . . . . . . .  2 1/2" !!!   Coincidence?  I think not. It would be unreasonable to employ the same bend diameter for significantly different conditions.


Is it your belief that AWS does not intend consistency between methods prescribed in D1.2 and B2.1 ?
Parent - - By electrode (***) Date 07-04-2013 18:05
Sir,

an interesting and most informative thread. Confirming your statement that says: "There is always something to learn here in the forum".
I seem to understand you, amongst others and among others, as being most experienced in correctly applying the codes.
And I also seem to understand, that, particularly in the US, codes seem to be the fundamental driving forces behind any welding action.

Please consider the following as a friendly comment rather than questioning the topic itself.

If I do understand that correctly the original post was, amongst others, asking for advice on eliminating bend test failure with 6110-T6 parent metal employing AA 4043 filler metal.
I myself was suffering similar issues as the OP a few years ago, welding however, AA 6082-T6 and employing AA 5083 (code requirement).

The parent plate did show 20.0 mm thickness machined for achieving a V-Bevel joint configuration (70° included angle).
One half of the parts welded were rolled structural sections that showed some special "backing bar" geometry to allow for semi-automatic welding and higher weld performance/deposition rates. In accord with the code (European Railway Vehicle Manufacturing) to comply with, we were preheating the parts to 120°C ± 20°C (~ 250°F ± 70°F), maintaining interpass temperatures 110°C ± 20°C.

GMAW-P (pure Argon) was process and arc mode chosen leading to, forgive me using this non-standard term/superlative, excellent results throughout in radiography and transversal tensile tests (acc. to EN 895).

The code (EN 910) though is likewise stringently requiring face bend testing (as-welded condition) and to keep it short, in order to avoid being judged again as writing "long-winded", we did consistently fail in trying to attain proper bending angles. On the contrary, as the OP, we did suffer failure most regularly induced at the weld toe.

Cross-check welds employing AA 4043 were providing us with similar behaviour/results, i.e. failure initiated most regularly at the weld toe area.

So, it was extraordinarily interesting to me reading 'bozaktwo1's' advice on the plunger type bending jigs (which we did use those days) and his recommendation on the rotary bender. Interesting, to say the very least. Maybe that could have been the root cause and the reason that we could not get the issues fixed finally, even after involving some external (famous) welding and joining institution.

Anyway, what I do honestly not understand through this thread is as follows. The OP speaks of issues also in meeting the tensile strength results required. And here it goes. Now, there are 2 different assessments. 'bozaktwo1' states that the OP is "going to have trouble getting those tensiles with 4043".  'tom cooper' whereas is saying "You should easily achieve 24ksi tensiles with 4043 as long as you don't char the he$$ out of it."

And, coming from that part of the world where the people are yet going on to sleep well (you can see me smiling while writing this) thus most likely being the reason making it hardly understandable for me, 'tom cooper' is also recommending to "be sure you (or your lab) anneals the coupons properly". And, in a second post, referring this advice to some code even.

Recalling the suffer I myself went through at that time I would have considered myself lucky having such valuable advice (especially in terms of the bending jig).

However, the (European) code would definitely not have permitted annealing the specimens for meeting the bending angle requirements – no way. It is surprising thus that the US code(s) seem to allow such procedures in order to jump the – obviously globally well-known – hurdles in conjunction with higher-strength aluminium alloy weld joint bending.

Finally it would be interesting to learn what your personal experience says when it comes to achieving or not achieving the tensile test requirements using AA 4043.

Thank you.
Parent - - By Lawrence (*****) Date 07-06-2013 16:28
Great thread guys!

Bravo
Parent - By electrode (***) Date 07-07-2013 11:43
Yes Sir.

That it truly is.

By the way and before I forget. Congratulations and good luck with your new job!

Thank you.
- - By 803056 (*****) Date 07-05-2013 02:17 Edited 07-05-2013 02:30
6061-T6 is often the alloy of choice because it has excellent as wrought tensile strength and yield strength. Elongation, not so good when it is in the T6 temper or the as welded condition (when compared to carbon steels). The two common filler metals are 4043 and 5356. If I were to make a guess, I would say the majority of aluminum alloy that is welded for structural application is probably 6061-T6 and it is welded with either 4043 or 5356. So, for me, this is an interesting discussion.

As noted, the wrap around bending fixture solves many of the problems experienced when performing the guided bend tests. As observed by a previous poster, the bends will often fail when using a plunger and die set because the HAZ is softer than the unaffected base metal or the weld, thus the bend is concentrated in the HAZ, and the bends fail as a result. The wrap-around bend fixture eliminates that problem. Another common mistake is failing to reduce the bend test samples to 1/8 inch when testing in the as welded condition. This requirement is included in AWS D1.2, B2.1, and ASME Section IX.

As for consistency between the various AWS welding standards, both AWS D1.2:2008 and AWS B2.1:2009 list different bend radii for M-23 base metals in the as welded condition and PWHT (annealed) condition. In the as welded condition both welding standards require the guided bend sample to be reduced to 1/8 inch (or less for job thicknesses). The bend radius for the 1/8 inch thick as welded sample is 2 1/16 inch (16 1/2 T for thicknesses less than 1/8 inch) in both AWS publications. Both AWS publication require full thickness guided bend samples up to 3/8 inch thick and both require 2 1/2 inch bend radius when the sample is in the annealed condition. Again, just so as not to lose track, we are limiting this conversation to M-23 (6061-T6 aluminum alloy). If the aluminum alloy is welded with an F-23 filler metal, it is treated in the same manner as 6061-T6 (M-23) in the as welded condition.   

I find it interesting to note that in AWS D1.2:2008 clause 3.12.2 states that PWHT must be addressed by the WPS. Clause 3.12.1 states that the PQR must record the essential variable and references Annex F for a sample form. Yet, I find Table 3.1 does not include PWHT as an essential variable. I did find an interesting bit of information in clause 3.8.1.7 Special Bending Conditions - M23, M24, M27 Base Metals, and F23 Filler Metals. It does include provisions for annealing the guided bend samples before bending them. Whereas D1.2 does not list PWHT as an essential variable in Table 3.1, I must agree with Tom that annealing the guided bend sample is permitted for the base metals listed  AWS D1.2.

Now let's switch gears for a moment and go back to AWS B2.1 which was referenced by the individual that started this thread. Clause 4.14 lists all the essential variables that apply for various welding processes, base metal groupings, i.e., M numbers, etc. Under the heading of PWHT, item 4.14.6 (2)  PWHT is an essential variable for all welding processes. So, I guess there are some inconsistencies between the various AWS welding standards. In the case of AWS B2.1 I would conclude that annealing the guided bend samples is not permitted for a WPS qualified in the as welded condition. The only time AWS B2.1 permit the guided bend sample to be annealed is if the WPS is qualified in the annealed condition.

This has been a great exercise. Tom, I agree with you that if you are welding in accordance with AWS D1.2 you can anneal the guided bend samples before performing the bend tests. However, in this case, whereas AWS B2.1 was referenced by the individual initiating this post, I conclude you cannot anneal the guided bend samples if the WPS is qualified in the as welded condition. I guess one really has to study the applicable welding standard before taking a position. Once again, I have learned something new. Consistency between all the AWS standards? Not from what I just read.

There are techniques involved or should I say procedures that must be followed explicitly in order to qualify M-23 aluminum alloys or alloys welded with F-23 aluminum filler metals. Short cuts always prove to be fatal. Aluminum is one of my favorite metals. I make a good living working with contractors that must qualify their aluminum WPSs. Some chose to ignore my advice and learn the hard way, others ask me what steps must be taken and we pass the tests the first time most of the time. I have had very good results qualifying aluminum alloys, even the M-23s, over the years. I spent nearly five years working on the Aegis missile system as a welding consultant for several of the main contractors working the project. It is safe to say I learn a lot about welding aluminum alloys during that time. If it was easy, they would not need my expertise.

Tom, you did good job of standing your ground. Still, how are you annealing your samples? Are you reducing the sample thickness to 1/8 inch or 3/8 inch? And what bend radius are you using? :wink:

Best regards - Al
Parent - - By tom cooper (**) Date 07-05-2013 10:07 Edited 07-05-2013 11:52
Hello Al-
To answer your question on my specimen preparation, to date I have only used 3/8" so the bend tests are always based on the 3/8 specimen configuration criteria; it would first be cut out of the plate oversize (water jetted, typically) then oven annealed ( as discussed, according to the recipe in D1.2) and then ground to final size; the mandrel diameter is as specified, 2.5-inches. I should also add that I don't personally do any of this but provide this direction to the lab.

I suppose we're at an impasse at the intent of B2.1 regarding the allowance of annealing M23 bend specimens.  My suggestion is to examine your conscience and tell me if you have EVER seen an application where the engineer intentionally PWHT annealed the 6061-T6 weldment.  It just doesn't happen and AWS would not have need to address that specific and bizarre situation.  If they wanted to address that bizarre situation, then they forgot to tell us what tensile requirement to meet because annealed 6061-T6 would surely NOT meet 24ksi.   Oh, and one more thing, annealed 6061 would no longer be an M23 alloy.

By my reading, AWS' intent is clear and consistent - bend specimens for M23 may be annealed whether we're working within D1.2 or B2.1; I don't have a copy of D17.1 handy but would be willing to bet it allows for the "annealed" modifier on bend tests also. 

Well this was fun, thanks for the challenge Al. I'll continue to anneal my 3/8" thick M23 bend specimens because AWS allows me to and suggest others also take advantage of their wisdom until a technical inquiry is submitted to the committee and a formal interpretation is disseminated.

Best of luck.
Parent - By 803056 (*****) Date 07-06-2013 02:40 Edited 07-06-2013 19:31
Hello Tom;

I don't disagree with you position that D1.2 allows the bend specimens to be annealed before performing the guided bend tests. I understand that the annealing operation makes the hardness and ductility uniform, i.e., there is no region that is harder or stiffer than another, so completing the guided bend test in a standard plunger and die bend apparatus is no longer a problem. However, the problems of bending the specimen are non-issues if the wrap around bending fixture is used. AWS D1.2 allows annealing, so by all means take advantage of the weasel words and use them to your advantage. No harm, no foul.

However, I look at each welding standard as a separate entity that stands on its own merits. Different committees are composed of different volunteers and each committee has its own ideas and beliefs of how a WPS should be qualified. Some of the reasons for the differences are perfectly logical; they are used for different applications. As it turns out, the B2.1 committee has a lot of influence from ASME Section IX. If you compare B2.1 and ASME Section IX you will notice there is more agreement between them than if you compare any of the D1.X codes to ASME Section IX.

There are times when logic cannot be the basis of determining how the WPS is qualified. As the inspector, we follow the code. When we don't understand how or what the code requires, we defer to the customer or the engineer. If there is an issue that cannot be resolved it must be referred to the code committee for an interpretation. I have submitted inquiries for an official interpretation on several occasions. I have prevailed in each and there have been changes in the text of D1.1 as a result of my inquiries. The problem with official inquiries is that they take forever. I have one that has been in committee over a year and one that has only now begun winding its way through the process.

AWS D1.2 does not include PWHT as an essential variable, whereas AWS B2.1 does. They are two different welding standards with different essential variables. Each has to be viewed and implemented separate from the other. It is no different if we were to compare AWS D1.1 and ASME Section IX. They are nearly different as apples and oranges. What is permitted by one is not permitted by the other. Trying to force fit one to meet the other without forethought before qualifying the WPS will only turn into a disaster.

As for your question; when would anyone want to anneal the 6061 weldment as part of the design? I have encountered that very situation in the aerospace industry where the members were welded, the assembly was annealed, then cold formed, and then reheat treated to maximize the mechanical properties. The final heat treatment after forming was a reheat to dissolve the alloying constituents, followed by a quench and artificial aging to recover the mechanical properties. They were not required to go back to the T6 condition, but I believe they did get back to the T5 condition. In that case, the client qualified the WPS in accordance with B2.1.

In the case of welding the M23 base metals; the problem is over-aging. Over-aging allow the alloying constituents to precipitate (out of solution) to the grain boundaries thereby weakening the material. That is basically what happens in the HAZ when you weld 6061-T6 aluminum alloy without controlling the interpass temperature. One way of mitigating the loss of mechanical properties is to rapidly cool the 6061 aluminum between each pass to minimize the width of the HAZ. That can be accomplished by the use of heat sinks, blowing cold air on the part between weld beads, applying wet rags to the parts between weld beads, or simply allowing the parts to cool between weld beads to minimize the interpass temperature. The issue is time at temperature. Anything that will minimize the time at temperature will be beneficial.

In the case of structural aluminum that is fabricated and welded in accordance with AWS D1.2; PWHT isn’t something that is commonly required. The code does have provisions for PWHT, but the Engineer must specify what the PWHT is and how it is to be accomplished. Personally, I believe it is a horrible oversight to omit PWHT as an essential variable, but I’m not on that committee.

I admit to being somewhat conservative in my applications of the different welding codes. However, being cautious has proven to be the safe approach over the years and it has kept my clients out of trouble. I try not to comingle the provisions of two different codes even if they are similar. This has proven over and over again to work for the benefit of my customers. Let me give you an example; when you qualify a WPS per D1.1, do you have to qualify fillet welds separately from groove welds? Now look at D1.6 and answer the same question; do you have to qualify fillet welds separately from groove welds? That is, in each case, are fillets automatically qualified if you qualify the WPS using a grooved plate?

By the way, this has been great discussion Tom. I enjoyed this conversation immensely. Which one of us is right? I am not sure it is a case of one being right and one being wrong. In the event of litigation, when the code is ambiguous, the court almost always favors most liberal interpretation of the requirements. The secret is to make sure there is no reason to litigate. Belts and suspenders work for me.

Best regards - Al
Up Topic American Welding Society Services / Certifications / Aluminum 6160-T6

Powered by mwForum 2.29.2 © 1999-2013 Markus Wichitill