Not logged inAmerican Welding Society Forum
Forum AWS Website Help Search Login
Up Topic Welding Industry / General Welding Discussion / 6061-T6 Aluminum Weld Crack Repair
- - By designrider (*) Date 02-25-2009 00:16
Longitudinal cracks in 3/16" fillet welds using 5356 filler wire on 6061-T6 base metal.  Inspection shows cracks most likely due to  inadequate base metal preparation and preheating.  I am the engineer, not the fabricator.  AWS D1.2 section 4.20 shows that the welds can either be repaired or removed and replaced.  Removing the entire weld would be a cumbersome project for the amount of welds that are cracked.  I assume removal of the "unacceptable portion" (ie. the lenth of the crack + 1/8" or so) is likely what needs to take place.  Does anyone have suggestions for repair?  Any help is appreciated.
Parent - By 803056 (*****) Date 02-25-2009 03:54
Refer to D1.2 for crack removal and repair and don't preheat the 6061 alloy. It is time at temperature sensitive. The mechanical properties can be degraded by over aging if the interpass temperature isn't limited.

I would excavate at least 1/2 to 1 inch beyond the detectable crack tip to ensure any damaged material beyond the detectable crack tip is removed. D1.2 includes the use of penetrant to determine the extent of the crack.

I typically cool the weldment using a fan or blower between weld passes and I never use preheat with that alloy.

Best regards - Al
Parent - - By Kix (****) Date 02-25-2009 13:25
    How thick is the 6061-T6 you are welding on?  Are you using the GTAW or GMAW process?  Have you determined why you are cracking?  The way you said you would do the repair is about right, but I would go out a bit further on the ends like Al said.  Hog er out, then penetrant test the area to make sure you got all of the crack.  Not to get off topic here, but some amount of preheat is good on thicker sections because hydrogen porosity is a common find inside fillet welds when using the GMAW process.  To high of interpass temps and time at temp is what kills the T6.
Parent - - By designrider (*) Date 02-25-2009 18:06
The base metal sections that show the most cracking are where 3/8" web stiffener plates are welded inside a C12x10.4 (which has a web thickness of 0.51") The fabrication used a GMAW process which I know produces more heat.  I was not present for the fabrication, so I do not know if preheat was or was not used.

 
Parent - By Kix (****) Date 02-26-2009 13:30
GTAW produces a wider HAZ then GMAW.  If your travel speeds are correct, you're not going to put that much heat into the base material as you would with GTAW.  I would not preheat over 250deg F.  I think 300 deg F is safe, but I still go with 250 deg F on 1/2" thick sections.
Parent - - By js55 (*****) Date 02-25-2009 15:51
How is it inspection shows its most likely inadequate base metal prep and preheating?
Parent - - By designrider (*) Date 02-25-2009 18:08
Base metal preparation was in suspect because some welds show a complete crack and detachment from the base metal of the thicker part joined along the side of the weld leg.  Most cracks are hairline fractures 1/8-1/4" long at the weld ends and are propagating longitudinally down the center (throat) of the weld.  However, I am open for comment on alternative origins of this problem.
Parent - - By aevald (*****) Date 02-25-2009 20:51
Hello designrider, along the lines of proper weld termination, ie. crater fill, if the welding equipment doesn't have provisions for settings parameters to accomodate this, you may consider making a suggestion to start the weld wherever it is designated to start and terminate short of where it is supposed to end and then weld from the end where it was supposed to stop back onto the other termination. In other words, for a given weld length, the termination will be back over the top of a portion of the bead and if it is a stitch it won't have an exposed crater at it's end instead it will have two starts with the termination somewhere along the length of the weld stitch. I hope you can visualize what I am talking about. Best regards, Allan
Parent - - By designrider (*) Date 02-25-2009 21:35
Good suggestion.  However, the visible cracks are on both ends.  To support your argument, there are several craters that exist at the end of welds on this project; these will also need to be repaired.  However, from what I have read, craters typically form star or x shaped cracks that can propagate into longitudinal cracks over time.  However, we already have very clear and evident longitudinal cracks at both the start and terminating ends of multiple welds.  If the weld starts are also a problem, then I fear this may not resolve the problem.
Parent - By aevald (*****) Date 02-26-2009 00:19
Hello again designrider, that being the case, you should consider fredspoppy's suggestion to use the 4043 filler. Another consideration if using the 4043 over the 5356 would be whether the parts will be anodized afterwards. If that is the case, you wouldn't want to use the 4043 as the weld deposits would turn black.  If the machine parameters are set up properly preheat should not be required. Best regards, Allan
Parent - - By Fredspoppy (**) Date 02-25-2009 16:59
You may want to select 4043 for welding 6061-T6 next time as it is known for being less crack sensitive for that application.
Parent - By Lawrence (*****) Date 02-25-2009 19:19 Edited 02-25-2009 19:21
I don't agree with somebodys statement above, that GMAW is a hotter process than GTAW on aluminum when you relate it to total heat input..  GMAW provides a high current at the arc, however travel speed may be increased more than 75% over that of GMAW so overall GMAW provides less heat input, a smaller heat affected zone than GMAW which will increase the chance of retaining more of the Heat treatment properties that your T6 condition provides.  Distortion control is also improved with GMAW.

Along with Fred's suggestion

Using a GMAW power supply that has adjustable controls for the weld terminations might be a solution.

GMAW aluminum spray transfer is sometimes difficult to control at weld terminations.. If the end of the weld is starved of filler or too much heat builds at the end of a thin section, the result if often cracking in the crater left at the weld termination. (which sounds like what your describing)

It's almost always about process control... 

Lincoln electric explains their newest crater fill function as follows:
http://content.lincolnelectric.com/pdfs/products/literature/nx290.pdf
The crater fill function
works similar to the start function, but in reverse.
Therefore, the starting WFS and trim would ramp
down from the welding WFS and trim over a preset
time after the operator releases the gun trigger.
Now with the TOTAL s2f(TM) welding process, the crater
fill function consists of two steps. The WFS and trim
will ramp down to a selected point for half the time set
for crater fill and then will stay at those values for the
remaining portion of the weld, called the cap. This
gives an enhanced and more consistent crater fill
than with previous methods and creates a button of
weld metal.
Parent - - By designrider (*) Date 02-25-2009 19:16
Also, If I (as the engineer) determine that some of the cracked welds are not needed for structural use, can I allow some of the cracked welds to remain unrepaired, assuming they have become nothing more than 'debris' or splatter on the frame?  Or will the existence of the cracked welds alone produce undesirable results (ie. in the exterior exposure, water leaks in at the weld cracks, freezes, expands, and puts unexpected strain on remaining uncracked welds). 

Thanks for everyone who has posted to this topic! Your comments are well received!
Parent - - By Lawrence (*****) Date 02-25-2009 19:25
Thats not a road I would want to go down.

You (as the engineer) would be ultimately libable if those cracks propigated to cause a failure.

The code requires the absence of cracks for very sound reasons.
Parent - - By designrider (*) Date 02-25-2009 21:23
Sound advice. I instinctively had similar feelings, but wanted to make sure I wasn't being excessive.
Parent - - By 803056 (*****) Date 02-26-2009 12:57 Edited 02-26-2009 13:00
When welding using the GMAW process, you need to verify the welders are in the spray transfer mode. Aluminum requires high heat due to its high thermal conductivity. Cleaning just before welding is very important as well. Aluminum oxide melts at a temperature about three times the melting temperature of aluminum (3600F vs 1200 F). Any oxide will inhibit coelesence and fusion type discontinuities will result. 90% of the maximum oxide layer forms in the first 24 hours after cleaning, so, cleaning the parts the day before welding isn't very effective. The quality level will dictate the extent of cleaning required and the means of removing the initial oxide. Chemical cleaning by dipping the parts into buffered sodium hydroxide followed by a clear water rinse followed by dipping in buffered nitric acid followed once again by a clear water rinse is very effective. Standing water must be removed by rotating the parts or use of a blower. Do not use compressed shop air! Still, wire brushing just before welding is necessary unless the welding is performed within about a four hour window.

When cleaning the aluminum use a clean stainless steel wire brush. Cleaning has to be aggressive in order to break through the oxide. It is very hard, that is why aluminum oxide is used for grind wheels and "sand" paper. Do not use a power brush, it will only burnish the aluminum and make welding more difficult. Grinding should be performed with a silicon carbide disk.

Welding progression should be "forehand", i.e., the gun is "pushed" in the direction of travel. As mentioned already, crater fill is important, so at the end of the travel the welder should reverse direction and pull the trigger in a couple of short "bursts" to fill the weld crater. It is common for the start of the weld to be "cold" and excessive convexity and overlap will occur at the weld "starts". Depending on the quality level required, the starts may have to be removed for the first 1/2 inch or so.

As mentioned, 4043 may be an advantage due to improved ductility in comparison to the 5356. However, the material thicknesses you mentioned are relatively thin and should accommodate any shrinkage that occurs during the solidification process. Preheating is not necessary if the welders are in the spray transfer mode for the GMAW. Check the weld size produced and the shape of the weld face. Concave fillets are more prone to "center-line" longitudinal cracks than convex fillet welds.

Photographs of the cracks may also provide us with more information than a text description. In the past what have been described as cracks have proven to be incomplete fusion.  Welding terminology is often a stumbling block to effective communication. I just consulted on an aluminum aerospace application this past week. The problems that were described over the phone proved to be a problem in terminology not in the welder's skill. What was described as cracks and "lack of fusion" were not the problem. Part of the problem was in sample preparation and interpretation of the micrographs. The other problems they were experiencing were related to the use of preheat, poor weld joint design, and the availability  of filler metal diameter that was too large for the application.

Best regards - Al
Parent - - By designrider (*) Date 02-26-2009 22:55 Edited 02-26-2009 23:02
Al;
This is very detailed advice and it is much appreciated.  You mention "depending on weld quality level required, the starts may have to be removed for the first 1/2 inch or so." Is this typical? Would the welder actually start the weld 1/2" long expecting to remove it later, or do I (as engineer) permit the loss of the 1/2" of specified weld length?  Also, in a previous post you said you would "excavate at least 1/2"-1" beyond the detectable crack tip".  Is there any significance to removing more than 1/2"?  If not, would it be acceptable to allow the removed cracked weld + 1/2" beyond (assuming it was a very small crack) to remain without rewelding? (i.e. in my mind this becomes equivalent to the idea of removing the first 1/2" of weld start as you said). I know welding reduces the strength of the base aluminum, but does rewelding reduce it further?  (From a structural perspective, welded aluminum takes a ~40% loss in strength, but no specification is given for a re-welded piece.  I assume the original reduction is conservative enough to account for this).  Regardless, if strength is further reduced by rewelding, it seems like rewelding should be avoided if possible.

On a different note, I have been advised in the past that aluminum works best in exterior exposure if it is welded "all around" joints.  I assume this is relative to water penetration at any unsealed joints (then potential freeze/thaw).  This raises the question:  Would allowing a weld portion to be removed and not replaced void the seal of the system...or is the concept above not really that important to maintain?

Thanks again! - Garret
Parent - By 803056 (*****) Date 02-27-2009 21:02
Sorry for the delay. My web browser had a glitch and I couldn't access the web.

Different welding standards have different acceptance criteria, therefore there are different quality levels that can be met. The highest quality requires close attention to every detail.

Many of the questions you ask are of an engineering nature. You, as the designer, know more about your requirements than anyone else. I have no idea of what you are building or what it's use will be. I'm not ducking the questions, just trying to be realistic and not lead you in the wrong direction.

Regarding how far past the crack tip is advisable; it depends on the alloys of the base metal and the filler metal, the thickness of the parts, rigidity of the assembly, etc. The best advice I can provide is to recommend taking a couple of samples with cracks and perform a macroscopic examination to determine how far the crack extends past the detectable crack tip. The crack usually extends a considerable distance when viewed under "high power" magnification. That will provide you with some concrete information that you can use to prescribe how far to excavate to ensure complete removal of the visible crack and the the crack that isn't yet visible. The use of penetrant will not provide the final word because even with PT, the crack has to be open to allow the penetrant fluid to be drawn in by capillary action. Likewise, the stain or indication has to be large enough and brilliant enough that you can see it.

There are situations where the in-house practices require the welders to make their welds "X" inches longer than the weld length specified to ensure the resulting weld is of sufficient length to carry the imposed loads. The welding standards typically assume the weld length being measured meets all the applicable acceptance criteria for profile, fusion, etc. Depending on the anticipated use, the weld starts may have to be excavated to satisfy customer requirements and expectations. Considering we're talking about aluminum, excavating the weld starts isn't an enormous amount of work. A rotary file (carbide burr) will make quick work of it.

As for water or fluid tightness; that is an engineering decision based on how the assembly is to be used and it's operating environment.

Best regards - Al
Up Topic Welding Industry / General Welding Discussion / 6061-T6 Aluminum Weld Crack Repair

Powered by mwForum 2.29.2 © 1999-2013 Markus Wichitill