Problem passing tensile test for AA6061-T651 & ER4043

Date 02-15-2007 06:35

You need to hear from the experts, but I will tell You what I know: The T651 heat treat properties will be lost in the HAZ, at least for a while. Some strength will come back in time. Have You achieved the 24 KSI reliably in the past?

By chall

Date 02-15-2007 13:15

A few years ago we spent a few months working with a number of aluminum base metals and both the GTAW and GMAW processes. What we discovered is that aluminum (more than any other metal I have worked with) looses tensile stregth very rapidly when it is "overheated".

Try it using a faster travel speed or by lowering your current. There will obviously be a trade off, but you should be able to get within spec, while still achieving an acceptable weld.

Charles.

ps - are you aware that ASME allows you go below the specified tensile stregth by 5%, provided the break is in the base metal and isn't the result of an obvious defect?

By js55

Date 02-15-2007 14:28

Question here. You mention the break is in the base metal. To specify, do you mean the HAZ? If so the certainly the heat regime is at issue. If not a more serious problem may exist.

By chall

Date 02-15-2007 15:20

I only have the 1998 edition of Section IX on my computer, so this needs verification.

QW153.1(d): If the specimen breaks in the base metal outside the weld or fusion line, the test shall be accepted as meeting the requirements, provided the strength is not more than 5% below the minimum specifed tensile strength of the base metal.

There is no requirement to identify the break location relative to the HAZ, but since the heating clearly has something to do with the reduced tensile strength following welding, it stands to reason that the failure will occur in the HAZ. (My opinion.)

Charles

By shaun420 (*)

Date 02-15-2007 22:52

I would like to thank everyone for their responses. I will keep all your advices in mind. Your help is greatly appreciated.

By 803056

Date 02-16-2007 02:42

Hello Shaun420;

6061 alloy aluminum is time at temperature sensitive. Holding it at an elevated temperature for a period of time will overage the material within the heat affected zone which can extend a considerable distance from the weld.

I have found that it is helpful to minimize the time at high intepass temperature (above 350 degrees F). I have used chills and fans to force cool the weldment between passes. Minimize the number of weld passes by increasing the tungsten diameter and welding amperage while maintaining your travel speed. Do not use weave beads. Maintain low interpass temperature (250 degrees F). That means you will have to cool the material between each weld pass.

Another issue that may result in low tensile strength is the fact that aluminum oxide has about the same density as the aluminum base metal. The oxides melt at more than twice the temperature the base metal melts (1200 degrees F for aluminum, 3200 degrees F for aluminum oxides). The oxides do not break down in the weld puddle and will not float to the surface. Instead the oxides can be entrapped in the weld deposit resulting in low tensile strength.

Use acetone followed by 90% isopropyl alocohol to remove any oils, grease, finger prints, etc. from the test coupons. After cleaning, handle the material with clean gloves. Likewise, clean the filler rods and do not handle them with bare hands. You can rub the rods with stainless steel wool to remove surface oxides. You can get stainless steel wool from most industrial supply houses. Do not use wool made from carbon steel.

Make sure the welder has removed all surface oxides using a wire brush with stainless steel bristles. Brush both the face and root sides adjacent to the weld groove in one direction, toward the weld preparation. Draw file (use a vixen or rasp file) the groove face as your last step. Do not let the welder check for burrs by running his or her finger across the cleaned material. Weld the joint imediately to minimize the reformation of the surface oxides. Wire brush between each weld pass just before depositing the next weld bead.

Remember to cool the test plates between each weld pass.

Good luck - Al

By MBSims

Date 02-16-2007 05:27 Edited 02-16-2007 05:29

I was surprised that someone else is going through the same troubles we are having with a 6061-T651 procedure qualification using GTAW and ER4043 at the same time. We have only managed to get the tensiles as high as 21,300 psi after considerable efforts to minimize heat input and improve cleanliness. We had some problems with a dirty puddle that seemed to be the result of a o-ring on the back cap not seated properly. Also noted higher gas flows (above 25 cfh) seemed to cause a dirtier puddle. Here's what we have tried so far:

1. 3/8" thick 6061-T6511 using 3/32" and 1/8" ER4043. All wire cleaned with scotchbrite pad followed by cloth wipe prior to use.
2. WP-26 torch with 5/32" pure tungsten, #8 cup, 100% argon at 20-25 cfh.
3. Miller Syncrowave power source using AC with continuous HF at 250-275 amps, wave balance varied from 30% to 50% to 65% on different test plates.
4. 70 F preheat, stringer beads only with as high a travel speed as practical, followed by immediate water quench to below 100 F.
5. Single V-groove with either 90 deg. and 120 deg. included angles with 0 root opening, backgouged and welded root pass.
6. Clamped test plates to 1-1/2" thick plate to dissipate heat during welding and control distortion.
7. Each bead cleaned with SS wire hand brush immediately before depositing next bead.

None of the 4 test plates sent out for tensiles has met the 24 ksi requirement. A couple of the test plates had welds that were so dirty (o-ring leak, possible dirty wire, gas flow too high) we did not feel confident in sending them out for tensile tests. All of the bend tests are passing. Tensiles are breaking in base metal HAZ. To be within 5% of 24 ksi, we would need a minimum of 22.8 ksi for a tensile failure in the base metal. My current opinion is that a 75%Ar/25%He mix is needed to support a higher travel speed. But, we would like to qualify this procedure with 100% argon if possible. If anyone has any suggested parameters bring the tensile results up, I would be interested in hearing them.

By DaveBoyer (*****)

Date 02-16-2007 06:30 Edited 02-16-2007 06:36

A marine enginering book I have lists 6061-T6 as 42,000 psi tensil strength in the unwelded condition and 22,000 psi in the welded condition. I think the specs for T6 and T651 are verry close, but I forget what the difference is. T6/T651 heat treat is achived by holding the alloy at elevated temperatures for a period of time, while at this temperature growth at the grain boundries causes strain hardening of the material if I remember corectly. The HAZ has been esentially annealed from the welding heat. That is why You have a ductile failure in the 6061-T651. How soon after welding is the test being done? A friend was developing the technology for producing the Corvett space frame from hydroformed 6061-T6 tubing. He mentioned that they could anneal the tube and hydroform it and within 24 hours it would be back to T6 properties. I don't know how much of the strength recovery was due to the strane from hydroforming and how much was from grain boundry growth over time. Grain boundry growth over time suposedly will recover some lost strength in a weld HAZ, but I dont know how much strength, or how much time.

By Lawrence (*****)

Date 02-16-2007 20:10

Shaun and MBSimms

I know what you are proposing *should* be doable.

None the less.

GTAW for 3/8 thickness Alum is on the outside of a reasonablly doable thickness range in my opinion. Especially in a production environment.

A change in process to GMAW/GMAW-P is going to increase travel speeds what, about 10X ? This will also improve mechanicals by virtue of a significantly slimmer HAZ. GMAW will reduce the HAZ and fill your bevel with fewer passes. And less attention to heat sinks and distortion control, A heat treatable filler may also help with dilution issues in the melt boundries.

The Notion of Ar/He if you must stick with GTAW is a good one.

Also... Loose the pure tungsten electrode. Go with Zirconium or Ceria. Put a blunted steep bevel for tip prep. This will focus your arc considerably and make for more control and quicker travel speeds.

Clamp your plates to copper rather than steel for heat sink.

The thing I like about GMAW for thick sections is that your Procedure Qualification run can mirror production with much more ease than can GTAW. There is less need for extreme heat sinks with GMAW and if you have GMAW-P you can weld fillets down to 0.060 with no burn through.

By 803056

Date 02-18-2007 03:39

Hello Marty;

You listed "ScotchBrite" for cleaning. I may be mistaken, but I believe the abrasive used with "ScotchBrite" is aluminum oxide. I would reconsider using any abrasive that utilizes the material, i.e., aluminum oxide, that I'm trying to remove. Use silicon carbide if you need to use an abrasive. Better yet, use carbide rotary files, vixen files, hand brushes, etc. to remove surface oxides from the base material and weld beads. Do not use power brushes. Power brushing will burnish the surface and drive the aluminum oxide into the surface of the aluminum.

Best of luck - Al

By MBSims

Date 02-18-2007 16:43

All good points. The scotchbrite seemed to be doing the job though. Test plates were prepped by grinding the bevels with new grinding wheels for nonferrous materials, followed by cleaning with new flapper disc, then SS hand wire brush (not a power brush). The last plate I ran the puddle was really clean, but it still did not pass tensiles. Generally, we would use GMAW for this application, but it is a maintenance project where GTAW was a better choice.

By Lawrence (*****)

Date 02-18-2007 22:35

Only one more thought.

Hot and fast!

With as large a filler rod as possible to cool the puddle.

By ssbn727 (*****)

Date 02-16-2007 18:34 Edited 02-16-2007 19:53

Hey shaun420!

Check out this web site and look over the topics... I'm sure you'll find something that maybe of interest for your situation... http://www.alcotec.com/techpage.htm

You can also get in touch with their Technical department and ask their experts what suggestions they may have for your situation... Here's the link for AloTech's Information Request page:
http://www.alcotec.com/request_info.cfm

Here's a different question with respect to weld cracking on 6061 T6:

Q: I am experiencing a weld cracking problem on our TIG (GTAW) production line where we weld thinner sections of 6xxx series aluminum metal sheets using aufhauser 4043 filler material. Why do you think my welds are cracking? And why is it that not all of my welds, but only some of them are cracking?

A: The aluminum/magnesium/silicon base alloys (6xxx series) are highly crack sensitive because they contain approximately 1 % Magnesium Silicide (Mg2Si), which falls close to the peak of the solidification crack sensitivity curve.

The Mg2Si content of these materials is the primary reason that there are no 6xxx series filler alloys made. The cracking tendency of these alloys is lowered to acceptable levels during arc welding by the dilution of the weld pool with excess magnesium (by use of the 5xxx series Al-Mg filler alloys) or excess silicon (by use of the 4xxx series Al-Si filler alloys).

When we TIG (GTAW) weld on thin material, it is often possible to produce a weld, particularly on corner joints, by melting both edges of the base material together without adding filler material. In the majority of arc welding applications with this base material, we must add filler material if we want to have consistently crack free welds. A possible exception would be counteracting the cracking mechanism by maintaining a compressive force on the parts during the welding operation, which requires specialized fabrication techniques and considerations. This method is seldom used.

I suspect that the welds in question that are not cracking are those that have had filler material added during welding. My advice would be to ensure that filler alloy is added to all welds during welding in order to reduce crack sensitivity. Consideration should also be given when evaluating the cause of cracking to any differences in welds associated with weld size, and variations in tensile stresses introduced by shrinkage, joint expansion, or externally applied loads.
Hope this helps.

Respectfully,
Henry

By MBSims

Date 02-18-2007 16:51

Henry,

A good link to the Alcotec website. I had gone thought my Reynolds aluminum welding book and done a similar search while this was going on. We also called the author of the Alcotec technical page (Tony Anderson) and picked his brain a bit for suggestions to get the tensiles up. His suggestions for our application were to keep the interpass temp down to 100 F max, rapid cool with forced air or water, keep travel speed as fast as possible, increase the current if needed to support higher travel speed, and use stringer beads only (no weave). He was very knowledgable and helpful.

By shaun420 (*)

Date 03-08-2007 00:29

Thanks all.

We have passed the tensile test with results of around 27 ksi.

We used a larger electrode and wire(1/8), made sure the material cooled to room temperature between each weld, and used a large steel plate as a heat sink.

We also used plate this time instead of flat bar. We were suspecting the quality of flat bar is not as consistent as plate.

Thanks again for all your help.

By DaveBoyer (*****)

Date 03-08-2007 04:38

There is a thread in the Shop Talk section posted by Kix that may interest You, particularly a graph posted by Lawrence.

By franciscocl7 (*)

Date 03-09-2007 20:11 Edited 03-09-2007 23:11

Shaun420

Simplemente estas afectando tu material tres veces, es comun que la fractura de tu material ocurre en la zona del metal base (zona sobreenvejecida -overaged), debido a que en es parte ocurre una transformacion metalurgica de los precipitados, los cuales esta transformandose de B¨ a B´ es decir que debido a la temperatura que se esta alcanzando en esa zona es suficente para que ocurra dicha transformacion. Cada vez que aportes calor a este material se va a ver afectado, y se va a ver en detrimento de tus propiedades mecanicas, debido al mecanismo de endurecimiento del mismo material. Nunca ocurre la fractura en la ZAT (HAZ) debido a que esta es una zona tratada por solucion, aunque los valores de dureza en esa zona no son muy altos. Tengo una grafica que obtuve de una union que hize con GMAW en transferencia por Spray. y todas las fallas ocurren fuera de la ZAT.

By MBSims

Date 03-09-2007 22:41

No intiendo

By franciscocl7 (*)

Date 03-10-2007 00:47

Simply, any time you heat input to material, the mechanical properties decrease due to the metalurgycal transformation of precipitates. they are longer in as weld condition than T6 condition.

By jerrykroll (**)

Date 03-09-2007 23:09

PLATE !!

Were you welding the "bar" parallel to the grain (direction of rolling) or perpendicular to the grain ?

This can make a HUGE difference in my experience. I have recent experience where toughness tests come out at over 50% less when performed
"with" vs. "against" the grain ! I am certain the same applies with tensile tests.

By aevald

Date 03-09-2007 23:20

Hello jerrykroll, I could definitely see where it would make a big difference as to the grain orientation, especially if the weld and haz were parallel to the grain. Tensile or break test, either one of them will find the weakest link and in this case, obviously that is in the haz area. Very good point. Regards, aevald

By franciscocl7 (*)

Date 03-10-2007 00:42 Edited 03-12-2007 15:36

Trhoug HAZ and beyond, the material is softness due heat input, tensil test specimens always break there. Measurements of hardnes from center of welt to the base metal have been demostrate that, mechanical properties decrease 75% of nominal rupture strenght.

ASTM B-557M
Al 6061-T6
1/4"
GMAW
1 pass.

By jerrykroll (**)

Date 03-10-2007 22:13

to franciscocl7 -

"Tensile test specimens ALWAYS break there"?

Tensile tests of welds can often break through the weld - one of the
biggest reasons that I perform tensiles on test welds !