liquid metal embrittlement from zinc-tin solders? 1

[HgTX]

Not sure if this maybe belongs in the coatings forum.

Anyway, LME due to tin in hot-dip galvanizing bath is well-established. But has anyone heard of cracks caused by tin in those solder "crayons" used for repair of galvanizing? And if so, did the problem go away if you switched to zinc-cadmium solders?

Hg

 

[rnd2]

Are these the so-called rub-on galv sticks?

 

[unclesyd]

I've never seen any problems with the sticks. Normally one wouldn't be at or near the temperatures long enough to cause LME. Also the normal galvanized steels are not prone to embrittlement.
In the US Cadmium is a no-no.
If you think there might some problem there are several good "cold galvanizing" materials available.

A little more details on the area of your concern would assist in giving a better answer.

 

[Rich2001]

After repairing galvanized structural steel (ASTM A572 gr 50) with the zinc-tin solder crayons I have not found any LME cracking using Magnetic Particle Inspection.

 

[hacksaw]

LME is what it says; there has to be a liquid phase in contact with the metal surface.

 

[HgTX]

Yes, these are the rub-on sticks, crayons, etc. Low-melting-point metal alloyed with zinc so that the solder will melt without melting the existing coating, or something like that.

Our supplier has a reproducible result of cracking when using the zinc-tin solders.

ASTM A 780 makes like zinc-cadmium solders are really common, but as unclesyd says, cadmium is a big no-no. I've only heard of one supplier so far.

What we're using this for is repair of galvanizing after application of ultrasonic impact treatment of welds (basically a form of peening). We don't like the so-called "cold galvanizing" (zinc-rich paint); we've found that it rusts through after only a couple of years--and that was at the higher zinc levels before ASTM A 780 lowered the requirements. Galvanizing after the UIT (a) undoes the effect (b) isn't an option in a field situation, and because of (a) we probably couldn't metallize either. The temperature of the solder doesn't seem to be enough to undo the UIT effect, but instead there are cracks.

Here's another question--does the tin cause the cracking, or does it just enlarge cracks that are already there but wouldn't have been visible?

Hg

 

[Rich2001]

Fracture can be initiated on a smooth surface without the assistance of any observable stress concentration, i.e.,a crack, though a stress concentrator increases the likelyhood of a major crack. For this reason, I have any repair Magnet Particle Tested after welding and prior to apply the crayons.

The grain-boundary local stresses generated by impinging slip bands can be identified with the initiation of microcracks. The atomic model is associated with a reduction of cohesive forces between the atoms of the solid metal by interaction with atoms of the liquid metal at the actual crack base. The SEM's that I had done on some galvanized induced cracks have shown tin at the crack tip, it is not clear whether the mobile atoms reach that point by surface diffusion or by vapor transport. The ASM Handbook mention that one model of LME suggest that vapor transport.

Here is a link to

http://www.scoss.org.uk/publications/

rtf/LMAC_Final_Version_2.pdf

 

[unclesyd]

What is the spec. for the material you are galvanizing?
Did you perform any testing prior to the stick galvanizing?

 

[Rich2001]

I didn't get to finish or edit my last post. (Wearing multiple hats... I had to handle a safety problem.)

Where is the cracking occuring, weld metal, base metal, both?
Depth of the crack?
Single crack or branched?
What flux are you using prior to stick galvanizing?

Did you remove the galvanizing before UIT?
The Gamma(75% Zn 25% Fe), Delta(90% Zn 10% Fe) and Zeta(94% Zn 6% Fe) are hard brittle layers, could you have cracked these layers forming stress concentrations. These has been some research on fatigue cracking starting in these layers propagating into the steel. Could the thermal stress induced when stick galvanizing be causing the problem?

 

[HgTX]

I'll have to get back to you on most of those question, but I'm pretty sure they did not remove the galvanizing before UIT.

Base metal can be ASTM A 572 Gr. 50, A 588, A 570 Gr. 50, A 595 Gr. A, or A 607 Gr. 50. I'll find out if any one of those was being used consistently.

Thanks everyone!

Hg

 

[HgTX]

Arrrhh!! Turns out UIT is a red herring. They do have problems with hot-stick galvanizing as described, but UIT isn't in the picture yet. We were looking at stick as a possible repair after UIT but so far they've used it for other purposes. I've asked all the other questions plus exactly what kind of damage they were repairing with the solder sticks. Will report back.

Thanks again.

Hg

 

[HgTX]

Me again...slight fix to Rich2001's link:

http://www.scoss.org.uk/publications/rtf/LMAC_Final_Version_2.pdf

or

http://tinyurl.com/5w438

I am researching the other questions, but maybe I need to put my initial question in a simpler way: Has anyone heard of these solders/sticks/crayons inducing any kind of cracking?

Hg

 

[rnd2]

I never have. Used them a few times over the years to coat galvanised pipe welds.

 

[HgTX]

A little more information (it trickles out slowly):

A welded piece came out of the (non-tin) hot-dip bath cracked. They removed the galvanizing over the welds, repaired the cracks, MTed, applied hot stick, cracks appeared when they got back after lunch. They removed the galvanizing from other parts of the piece where cracks had not been found, applied hot stick there as well, and cracks appeared.

Didn't happen to pieces that had come out of the bath okay.

Sounds to me like the tin solder (not to be confused with tin soldiers) was just enlarging pre-existing cracks that weren't showing up under MT, rather than creating them in sound material. IMNA metallurgist though.

What are the odds of my hypothesis vs. some kind of LME/LMA-style cracking induced by the solder?

Hg

 

[metengr]

HGTX;
I have been following your post. You have several different scenarios that need to be sorted out, and eliminated as root causes of your problem. For what its worth, I have the following suggestions;

1). Prior to hot dip galvanizing, perform a visual and wet fluorescent MT (WFMT) inspection of the component welds. This operation is inexpensive, and will determine if cracks are being introduced during production welding PRIOR to hot dip galvanizing.

2). If step 1) is performed, and you observe cracks, this indicates you are introducing pre-existing flaws into the hot dip galvanizing process.

3). If no cracks are found in step 1), proceed to hot dip galvanizing. If cracks are found AFTER hot dip galvanizing, have a metallurgical analysis performed to confirm if you indeed have LME. I don't know if you have this capability in-house, but you could send the defective section out for analysis.

3). If you are having LME, it can't be confirmed with just a visual examination or by WFMT. You need to have a metallographic examination of the defective area.

I know this sounds tedious, but it will help you to identify what is going on so that you can make process adjustments.

 

[HgTX]

It's not my process; I'm the end user of the product. And my real interest is the effect of the solder. I've specified it as a general repair method and now the galvanizer is telling me it causes the same kind of problems as those caused by tin in the hot-dip bath.

There's only so far the galvanizer is willing to go in this investigation. They didn't seem very excited about the cracks coming out of the bath; they get those every so often when the stars are aligned right, fix them, move on. Since this problem occurred on a piece that wasn't even our product, I can't make them do any destructive testing.

Whatever's causing the cracks in the bath shouldn't be LME. What's causing the cracks when the solder is applied...whatever that is, no one else seems to have come across it. The problem with cracking when there is tin in the bath is well established. If this solder caused the same kind of problem, wouldn't that be widespread as well?

Hg

 

[metengr]

I believe Rich2001 hit on some of the same questions I have that need to be addressed. If you are the end user, have one of the failed welds that had the tin-zinc solder applied examined to confirm the cause of cracking. At this point, all you are doing is speculating on causes.

Metallurgical analysis will confirm LME from the tin or some other problem associated with this process. My only experience with LME of tin on steel was during the time we had severely wiped (locally melted) babbit from a horizontal sleeve bearing on a steel shaft journal in service. I refused to believe it at first because the journal had developed pockets of spider web type cracking. We thought all along quench cracks from rapid frictional heating and subsequent cooling from oil until I decided to remove a ring sample because the shaft could not be repaired. I discovered the cracks were intergranular fissures caused by tin penetrating the grain boundaries. It was a hit or miss situation with these fissures.

I did find out that if you have spider type cracks on a steel journal after severely wiping a tin base babbit bearing, you must machine enough metal to completely remove any tin contamination. Otherwise, you will develop cracks in the repair area during welding, and will be chasing your tail.

 

[HgTX]

Thanks everyone for your help. I doubt I'll be able to get all the answers needed to properly answer the question. My main concern was whether I needed to dismiss tin solders as a feasible repair method (and whether the ASTM A 780 people need to worry about this), and based on the details I did get it seems to me like it was bringing out cracks that were already there but undetected.

Thanks again!

 

[HgTX]

Since I know y'all have been losing sleep over this, I figured I'd give you the fuller story I just got.

In a small subset of cases that have very high residual stresses (big thickness difference in joined parts meaning big weld to thin material) and stress-induced cracking, the residual stresses would be high enough to make the part susceptible to LME once the tin-containing crayon was applied. The stress cracks were small and well-behaved. The LME was pretty much exactly as metengr describes two posts above this one (9/14).

Most of the time the stress cracks are repaired no problem. Once or twice a year they'd get this problem where they'd repair the crack as usual, everything checks out with UT & MT, and within minutes of applying the hot crayons they'd have this spider cracking. When this finally happened in the presence of their metallurgist, they got some answers. They investigated and did find tin in the cracks, which had come from the solder crayon. Also the base metal, ASTM A 572 Gr. 65 Type 4, is particularly susceptible.

Looks like the reason no one's heard of this happening is it needs just the right combination of geometry and stress and material. It was happening for years at this manufacturer before they figured out what it was.

Not really what I wanted to hear, but there it is.

Hg