Welding Tips and Tricks

So my go-to rod for aluminum lately has been 5356 because I find the welds look nicer and it has higher strength than 4043. I recently built radiator end tanks for my old rad for fun and I happened to come across a 5356 spec sheet and saw that over 150F it is susceptible to corrosion stress cracking.

I did some googling and the default advice seems to be to use 4043 for heated applications. But the term corrosion stress cracking suggests that this is when the metal is in a corrosive environment under tensile load. The radiator stuff has nearly no load on it relative to the thickness of the metal, but it will live consistently at 200-220F.

Is this a real problem or is this for parts that are going to be loading the metal to a significant percentage of its yield strength while heated, or can it really crack at 200F with water inside it, while getting knocked about a bit (car vibrations)?

http://www.indalco.com/when-should-i-us ... -use-5356/

Hey D read this. This is the abbreviated version of longer articles but they all say the same thing about temps.

I've seen all of those but none of them define what they mean by corrosion stress cracking. What stress? Suspension component or a fluid container? I really doubt that boiling any joint of 5356 in water and holding it there a while, then gently pushing on it will make it crack - that's corrosion and stress above 150F. But I don't see that explained anywhere. I have seen anecdotal references to cooking containers and engine pistons fixed with 5356 without issue...

I am guessing the stress is the hot/cold cycling.

BigD wrote:I've seen all of those but none of them define what they mean by corrosion stress cracking. What stress? Suspension component or a fluid container? I really doubt that boiling any joint of 5356 in water and holding it there a while, then gently pushing on it will make it crack - that's corrosion and stress above 150F. But I don't see that explained anywhere. I have seen anecdotal references to cooking containers and engine pistons fixed with 5356 without issue...

'What stress?'

....uuummm.....pressurized rad and tank, cycling from maybe neg. PSI to 12#PSI or more, plus hot to cold cycling.
That kind of stress, which can be a combo of tensile, shear, twist, compression. Pressurized tank is akin to pressure vessel.
The metallurgical folks making this comment, positively had case studies to back it up.
They weren't trying to sell more 4043.
When the test lab guys recco sumthin', it helps to consider that; since the name of the game is to try and prevent potential problems--not create more of them, I think.
You can find anecdotal comments about using using 5356 for high temp app's.--whether or not it's believed to have 'worked'; doesn't negate test lab comments from a bunch of sources.

I understand they didn't make it up but I'm looking for specific metrics. Stress is not a measurement. 1 bar +- is certainly stress, so is a gentle touch, neither is a big deal to 1/8" aluminum normally. According to the metallurgists, some level is to 5356 above 150F... I would think what that is, would be available somewhere.

Just to clarify, I'm not looking to find out whether or not it's ok to use the rad, I always play it on the safe side when I don't know so I'm going to toss it. I just want to learn.

Also, I have some bungs welded on which I'm not sure which rod I used, can I just go over the welds with 4043 or cut out the weld and redo it?

Having a bit of success finding more technical articles incl research papers. The corrosion bit seems important. Eg weld in a boat hull with exhaust running along the weld.

When you weld and the whole reason for different alloys of filler is that when the metal enters its molten state, some metals and components will boil off, its just going to happen. Now your alloy has lost some of its components and its chemical balance is now, well, unbalanced.

The filler rods purpose in life is to make up for that as good as it can and provide the missing alloy with what it needs to stay together. So when you pick the filler rod best suited for joining 2 metals, its not like its the best type of glue but it will replenish some of what was lost during the welding keeping the metals as legit as they can be.

Its especially noticeable when doing stainless as it requires that special treatment (passivation) after most welding to remain corrosive free. That is due to contaminants and some metals floating to the top of the metal now breaching and preventing the oxide layer from regenerating. By passivating it you pull off all the iron particles (which are so small you won't ever see them) leaving only the chrome surface behind which can then reoxidize and passivate itself.

How I got here I am not really sure.. Oh right.

In regards to your welding over the bead, that's a no-no. Imagine you do that now all of the penetration you got with your 5356 that is buried in the metal cracks once enough stress and heat is applied, your 4043 weld will be much, much less strong if even strong at all as its not biting down on the parent metal but on the 5356 weld. The correct procedure when having used an insuitable filler rod is to cut it out and remove it to the best of your ability, some remains may get into the 4043 weld and that's pretty much inevitable in most cases but do your best to remove it and then reweld it.

Aluminum is odd, I built stuff with it a lot and I love it but it is an odd metal and its very different from any other in the way it reacts to welding, if you treat it right it will be nice to you. Don't give it what it needs and it will give you hell in terms of cracking. Like running a fine bead down and you can see the crack following your torch

There, first lesson of the day.

The reason I ask about a cover pass with 4043 is that according to what I've been reading, it's the magnesium deposits that are problematic when exposed to corrosion under stress. Furthermore, one paper I found showed that the failures in the study were limited almost exclusively to 7% content metals. 5356 is already 5% but I'm wondering if further dilution with 4043 would mitigate it even further, through both reduction of magnesium concentration and protection from corrosive elements (which for al-mg alloy is halides which are present in bodies of water like lakes and rivers - hence greatest concern being for boat applications).

BigD wrote:The reason I ask about a cover pass with 4043 is that according to what I've been reading, it's the magnesium deposits that are problematic when exposed to corrosion under stress. Furthermore, one paper I found showed that the failures in the study were limited almost exclusively to 7% content metals. 5356 is already 5% but I'm wondering if further dilution with 4043 would mitigate it even further, through both reduction of magnesium concentration and protection from corrosive elements (which for al-mg alloy is halides which are present in bodies of water like lakes and rivers - hence greatest concern being for boat applications).

I would assume it would. Grinding it out and rewelding with 4043 will probably be the easier way. At least you know you did the right thing based on what you know.

Yep, definitely the right way to go. I'm one of those guys that doesn't sleep at night knowing I didn't do something right, so it will be redone.

These may be interesting .

http://www.esabna.com/us/en/education/b ... -deg-f.cfm

http://www.esabna.com/us/en/education/b ... elding.cfm

Also there was a thread on here a while back on the welding of AN fittings. My .02c is you won't have any problems with your rad. Prior to 4643 being available, some teams would heat treat 6061 fluid lines to T 6 and these were welded with 5356.

I'm wondering what " prolonged elevated temperature " is in terms of time and temperature.

BigD wrote:Yep, definitely the right way to go. I'm one of those guys that doesn't sleep at night knowing I didn't do something right, so it will be redone.

Well, we have something in common as I am that way too. Sucks being a perfectionist sometimes but I wouldn't have it any other way.

BigD wrote:Yep, definitely the right way to go. I'm one of those guys that doesn't sleep at night knowing I didn't do something right, so it will be redone.

I am the exact same way

That's the main reason I went into learning about metals and metallurgy, I am not very good at it but I know enough to get me by knowing I did the right thing or how to fix what someone else messed up

kiwi2wheels wrote:This may be interesting .

http://www.esabna.com/us/en/education/b ... -deg-f.cfm

Also there was a thread on here a while back on the welding of AN fittings.

http://maxal.com/files/QuickSiteImages/ ... 12_doc.pdf

That is my bible for aluminum right there.

kiwi2wheels wrote:This may be interesting .

http://www.esabna.com/us/en/education/b ... -deg-f.cfm

Also there was a thread on here a while back on the welding of AN fittings.

That was mine too At least I'm finding out about these things before the parts go into service. Interestingly enough before cutting the AN fitting flange apart to take out what turned out to be 2024, I hammered the piss out of it and bent it to hell but it never cracked anywhere

BigD wrote:

kiwi2wheels wrote:This may be interesting .

http://www.esabna.com/us/en/education/b ... -deg-f.cfm

Also there was a thread on here a while back on the welding of AN fittings.

That was mine too At least I'm finding out about these things before the parts go into service. Interestingly enough before cutting the AN fitting flange apart to take out what turned out to be 2024, I hammered the piss out of it and bent it to hell but it never cracked anywhere

If I remember the AN thread, 5356 was the go to wire that was mentioned. I always used it, as I am not a fan of 4043; I like the ductility of 5356. Plus 5356 is used in repairing heads in the chambers and fire face.

4943 is also coming up in the ranks.

That was mine too At least I'm finding out about these things before the parts go into service. Interestingly enough before cutting the AN fitting flange apart to take out what turned out to be 2024, I hammered the piss out of it and bent it to hell but it never cracked anywhere[/quote]

Conversely--I've had aircraft sheet items of 2024 for crack repair and every time-HOT CRACK.
Told the a&p 'why?' this was occurring--only to hear 'Well--some places are welding them.'
He wouldn't stand for the proven reason why 2024 is not weldable.

Olivero wrote:

kiwi2wheels wrote:This may be interesting .

http://www.esabna.com/us/en/education/b ... -deg-f.cfm

Also there was a thread on here a while back on the welding of AN fittings.

http://maxal.com/files/QuickSiteImages/ ... 12_doc.pdf

That is my bible for aluminum right there.

Thanks for that. But they don't give the temp range for 4943.

dave powelson wrote:
Conversely--I've had aircraft sheet items of 2024 for crack repair and every time-HOT CRACK.
Told the a&p 'why?' this was occurring--only to hear 'Well--some places are welding them.'
He wouldn't stand for the proven reason why 2024 is not weldable.

I watched someone weld a 90 degree flange to a .040" 2024 T3 panel using strips cut from a sheet of the same material

Fortunately (or un...) it was for a car..................

Olivero wrote:4943 is also coming up in the ranks.

That's what I'm going to try next, sounds like the magic bullet - nice properties of 4043 and strength closer to 5356.

OK, not trying to jack this thread, but since discussion involves cooling system components I figured this could possibly help the OP out as well. (thanks Big D, you may have saved me some hassle)

Got this thing here tacked up and ready to weld.
Tee fitting there I made from 6061 t-6 and the tube is unfortunately an unknown grade of aluminum. What I do know is it comes from Japan, has a 1" OD with about a .045" wall an is pretty soft, much softer than 6061 in t-6 condition. In use it will mostly see 170 to 180 degree temps, with possible spikes to 210 to 215.

Question is, what filler to use? Plan was to 5356 it, but now this whole 150 degree max thing has spooked me.
I have both 1/16 4043 and 5356 on hand, but no 4943. Any suggestions?

dunkster you better ask Steve aka Otto...he's the encyclopedia.