weld failure

Date 10-31-2009 20:10 Edited 10-31-2009 20:40

Hi gonzo

Just to clarify: Your using 5356 for Filler? Need to know filler and base metal, also are those cast or machined flanges?

By gonzo

Date 10-31-2009 20:32

Good afternoon Thomas
Tube base metal 6061 the flanges same base metal machined
filler rod ER5356 x .045
the Drawing had no call out on what type of welding rod to use
everyhting looked ok welded leaked checked heat treated at 550f (just got them annealed )
final mechined and leaked tested again
this was done like 5 months ago so now the customer call .they leak
I'm just wandering if the weld prep and weld were to small
thanks for the reply Tommy

G.G

By Tommyjoking (****)

Date 10-31-2009 21:13

Good day!
I am not sure how well 5356 responds to pwht on 6061...but I understand the choice otherwise. It sounds like they were welded with sufficient amperage to negate the size of the joint prep. I am curious about the service this weldment is exposed to.....do you know if there are any temperature extremes (especially prolonged exposure to over 150F), extreme vibration ?? If they were installed and working fine then failed in service it might be a poor choice of design/alloys for the job. If you have any that have not been installed that you can look at, some magnaflux or xray might be a good idea just to cover your end of it. There are some very knowledgeable folks on here and I am sure some one could provide some definitive insight....sorry that is all I can think of at that moment. Look forward to hear what folks have to say bout this one.

Best regards
Tommy

By gonzo

Date 10-31-2009 22:23

hello Tommy
as far as the temps onthe tube none there water cooled the choise of rod was just
for the strength on the small weld prep (shoul've went with 4043 or 4643 )as far as the heat treating
I would believe it would just stress relive it ( I really don't know)
I assumed it wouldn't be enough for cracking the weld
thanks for the help
Gonzo

By bozaktwo1 (***)

Date 11-01-2009 13:14

Just to elaborate on Tommy's post. I do know how 5356 responds to heat treat. It doesn't respond well at all. If you're trying to reach a specific temper after welding, try 4043 filler.

By ssbn727 (*****)

Date 10-31-2009 22:27

Hi gonzo!

On the 6061 aluminum designation for the grade you're working with, does it also include a letter, followed by a number with a dash in between the letter and number starting form left to right and beginning just after the 6061 aluminum grade??? (e.g. 6061 T-6)

Respectfully,
Henry

By gonzo

Date 11-01-2009 21:03

I believe it was ( 6061 T-6 ) I'll have to look at the specs
of tube and flange
thank you
G.G

By gonzo

Date 11-04-2009 04:25

TommyJoseph,Henry,Al,Curt just wanted to say thank you to all ,for all of your help and for all the Information you
help others and my self on the AWS forum . on my problem with the 6016 T-6
we took the tubes to get leak checked by the people that certified them the first time

we're good ..the welding with the ER5356 at the 550F temperature wasn't the problem
it seems that when the customer gets the tubes they go out to get blasted for the process
they coat them with. They're sputtering targets ,so I guess the sand blasting being so coarse
created micro cracks on the welds now we're just going to make a bigger weld prep
to get 100% penetration and make the weld as thick as the tube wall to take the friction of the sand blasting
were going to remove the weld machine a weld prep (no coolant),flanges are tacked on the inside of tube
weld prep will be 45 degrees x 100 depth I'll try Dc with helium two passes

once again sir's thank you

gonzo

By ssbn727 (*****)

Date 11-04-2009 13:00

Hi Gonzo!

May I suggest that you practice with some scrap first in order to get used to GTA welding DCSP with helium, and check out the results you get before you work on production welding of these components??? Better to be safe than sorry!!! ;) ;) ;) The grit blasting tends to cause all sorts of issues... What is the blasting media being used in the blasting process???

Respectfully,
Henry

By gonzo

Date 11-04-2009 20:30

Good afternoon Mr Henry
I'll test on some samples first to get the DC helium dieled in
the tubes are getting blasted with Garnet crystal media either a 50 or 60 mesh
so now the customer wants to put a 50 degree (100 degree conbined ) chamfer x .100 deep to make
the weld 100% with a .050 max weld reinforcement so it makes the weld
about .240 wide x 100+deep I'll try it on some coupons to see what results I get
thank you for your advice Mr.Henry
gonzo

By Lawrence (*****)

Date 11-04-2009 20:57

Henry is right about doing some trials.

I suspect that control of melt-thru (root penetration) with DC- is going to be hard to keep consistant. DCEN is pretty good especially for fillets on thick (>1/8th) aluminum)

Do you agree Henry?

I know Joe Kane mentioned on the other thread on this specific problem that Alternating current produced welds with less density and DCEN producing welds that are more dense.. Joe I hope your following this because I don't quite Grock what you mean??

Anybody else have any ideas about weld density ... AC vs DC and alum GTAW.. Or how GMAW might compare to this question????

Very good thread.

By ssbn727 (*****)

Date 11-05-2009 01:59

Hi Gonzo, Larry and Joe as well if you're still following this thread ;)

Yes! I do agree Larry, and that's why I suggested to run some trials as you put it to see what sort of consistency Gonzo ends up with... what I wasn't sure of until now since he's elaborated further, was the fact that his company is now actually changing the joint geometry in order to achieve CJP (Complete Joint Penetration) as opposed to just thinking of the possibility to do so, and there's nothing wrong with doing so as long as everyone in the chain of command is on board especially the customer and their representatives...

Now with that being said, I believe the questions I'm about to ask you are warranted, because after all it will determine whether or not you have the capability to customize your weld penetration profile to where you and the customer desire it to be based on the configuration of your stated joint geometry as well as the relative difference in thickness the 2 parts have with each other (i.e. the tube wall and the flange itself).

1.) What type of power source are you using to weld these parts together, and do you have the capability really customize the waveform, i.e. true square wave capability as well as a very wide margin of pulse capability, i.e. low to very high hertz range to dial in for controlling not only your penetration, but also the weld profile as well??? I ask this because an inverter type power source with true square wave capability as well as a very wide frequency range for pulsing would certainly be a better choice when it comes to controlling the weld penetration profile consistency as Larry (Lawrence) has already alluded to. ;)

I also think that the differences in the thickness of the two parts would make it real hard to achieve similar consistency with this type of joint such as a flange to a tube as opposed to welding it instead with the type of power source I just mentioned including the application of pulsing with even straight argon alone... At the thicknesses you're welding together Gonzo, there really is no need to use helium since it is relatively easy to achieve CJP within those thicknesses, and adding helium may just make matters worse with respect to controlling the consistency of the weld penetration profile which would be of primary concern with this type of joint...

2.) Correct me if I'm wrong but, if you're going to weld the flange to the tube, you're describing it as the tube having a wall thickness of 1/8th of an inch, or .125 of an inch thick to a flange which has an overall thickness of .101 of an inch thick right? So, you're talking about a difference of .024, or 24 thousandths of an inch which is probably why in the drawing you posted, one member has a "J" groove geometry and if I'm correct, that is machined on the flange right? Unless the "J" groove is machined to the tube wall itself where the tube is cut to length which either way looks very unclear according to the drawing you provided in your earlier post since it doesn't show the differences in thickness between the two parts (tube and flange) when they are shown together in one view shown... I say this because what I'm interpreting in the drawing is two members of the same thickness being joined together instead of what you typed which would be the thickness dimensions I wrote in my first sentence of this paragraph...

In any event, the purpose of the "J' groove may just be incorporated into the joint configuration so that residual stresses in the weld can be minimized somewhat, and enough so that when the component is Post Weld Heat Treated, certain behaviors may be controlled or minimized in order to avoid some forms of cracking either in the weld or adjacent to it as well. ;)

It also doesn't show the orientation of where the weld is to be deposited in relation to where the flange meets the tube wall... In other words, you wrote: "They're sputtering targets ,so I guess the sand blasting being so coarse created micro cracks on the welds now we're just going to make a bigger weld prep to get 100% penetration and make the weld as thick as the tube wall to take the friction of the sand blasting were going to remove the weld machine a weld prep (no coolant),flanges are tacked on the inside of tube weld prep will be 45 degrees x 100 depth. I'll try Dc with helium two passes." I hope that I re-wrote that paragraph in the way you expressed it! ;)

Okay, now where I'm confused in this paragraph you wrote, is when you write: "remove the weld, machine a weld prep using no coolant..." Then this is where I'm a bit confused because of the drawing... "flanges are tacked on the inside of tube." Then you finish with: " weld prep will be 45 degrees x .100" depth... I'll try DC with helium two passes." Now did I interpret that in the manner which you meant ot describe what you guys wetre going to do next along with adding another 10 degrees included angle into the "V' groove geometry???

Also, if the tacks are to be placed inside the tube, does this mean that you are going to orient the inside surface of the tube parallel, or perpendicular to the face of the flange, because it really isn't clear in the drawing and would not make sense to me if you were to orient the face of the flange parallel to the inside surface of the tube wall... Now if it is the surface which represents the thickness of the flange is to be oriented with the inside surface of the tube wall, then it would make perfect sense, so that is where the drawing is very unclear in showing htis important factor I'm emphasizing so far... I hope you understand what I'm getting at so far - CAPECHE??? :)

Oh and by the way, Gonzo, you do not have to call me "Mr. Henry" even though I know you mean it as a sign of respect and I do appreciate it - really but, it's not necessary and you will not be disrespecting me by calling me by my first name only either - CAPECHE??? ;) ;) ;0 In fact, you can even call me "Hank" if you want to, okay??? :) :) :) So I'll await your clarification on this discrepancy although, please do not interpret this by stopping what you all plan on doing with respect to fabricating and welding together some trial sputtering targets because that's what you need to get ready - first and foremost, and remember that we're all just throwing out suggestions to you without any warranties or guarantees that what we suggest will indeed work - Okay??? Btw, I almost forgot to "Weldcome you to the World's Greatest Welding Forum!!!" So once again, "WELDCOME!!!" :) :) :)

Respectfully,
Henry

By Tommyjoking (****)

Date 11-05-2009 10:13 Edited 11-05-2009 10:59

Hi again Gonzo, thanks and your very welcome, no big deal I enjoy trying my best to help, that's pretty mediocre help but I try.

WOW!!! Henry calm down brother LOL!!! I was waiting for you to show up here!!!

Yep tacking inside is 100% necessary if you want to machine the new joint in one pass, otherwise the bevel will have to be carried out individually. Dry machining is a good idea, especially if its not a high production effort.....don't need any contaminants complicating your results. My personal opinion is you already achieved 100% pen on the first joint prep....the second configuration is not changing much.

Is the "microcracking" in the weld pool itself or on the toes or parent metal>?

I am very interested in the deposit density discussion conducting DCEN, over here Joe(!!!), but I feel the helium is a waste of resources here, straight argon will suffice I bet. Henry is 100% in the fact that you will need a quality powersupply to pull that off with any consistency...also it will have to be dialed in, test pieces are mandatory, per Law and Henry.

I also believe you do not need a DCEN process to achieve reliable results on this "particular" part. I think you need to consider a much more gentle alloy if you intend to stress relief with annealing,,,like 4043. If you can avoid the stress relief stay with the 5356...send it in an as welded condition.

I may be totally wrong here but this disturbs me:
"so I guess the sand blasting being so coarse
created micro cracks on the welds now we're just going to make a bigger weld prep
to get 100% penetration and make the weld as thick as the tube wall to take the friction of the sand blasting"

Gonzo if you are developing flaws in your joining process on a after treatment.....something is inherently wrong with the weld process to begin with ...the sandblasting is merely exposing pre-existing flaws in your weldment. The coating that is applied after blasting...is this a baked coating??? Between "annealing" at 550f with 5356 and any unknown customer imposed heating you may be headed for lots of cracking problems......I am no alloy expert whatsoever, that is why I defer to these guys ...but I see the potential for a bad combination of processes here and you need to ask the right questions to find the answers.

Having said my little piece not near so thoroughly as Hank, I will shut up and let the pros handle this because there are some issues.

Best of luck and best regards
Tommy

By gonzo

Date 11-05-2009 17:52

Good morning Sir's
Ok I'll try to clear this mess of confusion I created I'll start with an apology
as you guys can see I write as bad as I draw ...new drawing shows the shoulder on flange an
tube is machined to 1/8 wall at the flange area

the new weld prep is by customers engineers they want their tubes to be
able to pass leack check at the high vacuum specs
The media blasting is for their coating applications
it has to be rough in order for the coating to bond to tubes
they haven't privided any info on what process they use to bond the coating
material to tube waiting for response
(confidential info .....I just want to know if high temps are involved )
I'll run some test's with both 5356 and 4043 get them blasted and leack checked
parts are not going to annealing and welds are not getting machined of afterwards
they want .050 weld reinforcement
so I'll get everything tested and hope for the best
as for the micro cracks nothing after welding or annealing it happened after Garnet crystal blasting
I'm welding with a 90's syncrowave 350 haven't been able to talk them into buying a new one
ok sirs once again thank you for your patience and sorry for the confusion
it the side of my brain that thinks in spanish but has to write in English

gonzo

By Lawrence (*****)

Date 11-05-2009 18:55

Your old Synchrowave will do just fine for this kind of joint.

Set balence control around 7 of 10 (toward max penetration)

Boost filler wire diameter to .090 or 3/32 this will have a cooling effect on puddle.. (.063 absolute minumum)

I don't think 40 or 50 series fillers is related to your issues.. nor is color match so I think it's a "pick-em"

I have a hard time with the notion of post-weld blasting being a cause either. If it were pre-weld blasting than it might be an issue.

If you were doing a large production run of these thing-a-ma-bobs GMAWP (with hotstart and crater full options) might be more suitable.

Keep us informed!

By ssbn727 (*****)

Date 11-05-2009 20:17

"Oye Gonzo!"

"No te precupar tanto si usted esta pensando en Espanol!!!" However, most of in here speak English first so maybe we could stay with it as the language to communicate because I don't want you to think you did anything wrong buddy, it was just the drawing that had me a little confused because I couldn't see clearly how the 2 members were oriented as they were to be joined together - that's all. ;) Everything else you have described so far has been very good with respect to your own capability to interpret it in using English as the language of communication, so you're doing an excellent job so far!!! ;)

Now, with respect to the type of filler metal you are using... Does the customer insist on using a 5356, or are they open to an alternative such as a 4043, 4643 or a 4047 instead??? Look at this link for a description of an Aluminum 4047 filler metal:

http://www.esabna.com/us/en/education/knowledge/qa/What-are-the-advantages-of-4047-filler-alloy.cfm

Here are some more questions and answers on Aluminum from Tony Anderson who is one the top Aluminum welding experts:

http://www.esabna.com/us/en/education/knowledge/qa/

Here's another great bunch of links from AlcoTec:

http://www.alcotec.com/us/en/support/index.cfm

http://www.alcotec.com/us/en/support/Alloy-Data-Sheets.cfm

http://www.alcotec.com/us/en/support/Alloy-Selection-Chart.cfm

http://www.alcotec.com/us/en/support/industry-applications.cfm

And here's the link to ESAB's Education home page:

http://www.esabna.com/us/en/education/index.cfm

In other words, are the specifications written by a government agency of some sort??? because if it is, it may be hard to convince them change the type of filler metal without supplying evidence showing that the 5356 filler is doing more harm than good in this particular application... The customer needs to understand that they need to change the filler metal to one suggested by all of us instead for whatever reason that may be why they prefer the 5356 unless they have a convincing rationale for the use of the magnesium rich 5356 as opposed to a 4043.4643, or a 4047 which may just be the ticket for your application including the intended service application.

I mean, it reads a bit weird in that the welded targets have to pass helium leak testing and a 5356 filler is to be used since it is a magnesium rich filler as opposed to a silicon rich filler such as a 4043, 4643, or a 4047 instead which is recommended by almost all of the real experts I know of... Look Gonzo! Take a few minutes to read this link I'm providing you from Lincoln Electric who's name is known worldwide with the phrase: "The Welding Experts." You can read that perhaps the subsequent heat treatment which is being applied to the welded sputtering target is not the optimal one to use, and that further review with respect to the heat treating procedure your company is using could be something to consider... In any event, here's an article written by FranK Armao of Lincoln Electric which I think you will find rather interesting to say the least:

http://www.lincolnelectric.com/knowledge/articles/content/comistakes.asp

However, if it is not something that is written in the specification to switch to a 4043, 4643 or a 4047 filler, then go for it Gonzo!!! ;) Bozaktwo1 does make a very good point about how 5356 responds to heat treatment in that it doesn't respond very well in comparison to the 4043, 4643 or the 4047 filler instead.

I think you'll do fine with the Syncrowave you have currently available to work with as it is a good power source to weld aluminum with even though the newer ones are true square wave power sources in comparison... The Syncrowave does have some of the square wave characteristics however limited in comparison to an inverter type power source which enable you to do so much more that you cannot do with even such a fine machine as a syncrowave 350. However, it is a very good machine nonetheless for aluminum.

Well, I hope this post will help clarify some of hte issues a little bit better for you!!! "Entonces - Que Dios Te Bendigas Hermano!!!" So - May God Bless You brother!!! ;)

Respectfully,
Henry

By gonzo

Date 11-05-2009 21:03

thank you Henry
you just sold me on the 4047 but for rigth now I'll have to go
with the 4043 just cause the timing ,they need them yesterday
and I know I don't say it enough but I do have a lot
of respect for the knowledge you have and
for the help you guys provided here in the forum
just getting ready to weld 2 of the 8 tubes
thank you
gonzo

By ssbn727 (*****)

Date 11-05-2009 20:44

Hi Gonzo!

That is a much clearer drawing and it seems to make sense now why they want a "J" or a "U" groove for the weld/joint preparation since it has to do with minimizing residual stresses around the weld as Professor Thomas Eagar talks a lot about in his lecture series @ MIT which if you look at online, you'll understand exactly what he's referring to especially with thinner materials with respect to joint geometry and weldability issues such as distortion, warpage, residual stresses, shrinkage problems, delayed cracking issues etc.

I'll post the link later but, i've got to run before the post office closes, so I'll back shortly!!! :)

Respectfully,
Henry