Crackes in steel structure

[TeoAlfa]

Hello to everyone here!
We recently faced a problem at work with a complex steel structure. The structure comprises of a large and bulky piece of special steel component (i think its 4130) and quite a lot of S355 steel parts welded around it. Think of it as a large 4130 box with a number of S355 "legs" welded around it. The part had been welded but the specifications did not require any PWHT (point#1). The whole structure then was sent to be hot-dip gavanised. Don't ask me why, we simply followed the specs.
The structure after the galvanising created some large cracks. Both around the welds and also on the 4130 component itself!
I have to note here that the galvanisation temp was around 450deg Celsius and i have no idea hw much time the structure has been keeped inside the bath.
The question is: The problem arised due to no PWHT or because of the hot dip galvanising itself?
I strongly believe it's the first but i would like to read your comments though.
My view is that at 450deg Celsius the steel does not exceed any critical transoformation temp point so it would be impossible to state that galvanising caused any martensitic formation. Welding on the other side causes the metal to come to temps above Ac3, so afte welding thee was enough time (and cooling rate) to form some percentage of martensite.
Do you think it's anyway possible to form any martensite by solely heating the component to 450deg Celsius, and let's say cooling it at any rate after that? Anyway, the structure has been brought to 450deg, then cooled at the air.
Thanks for any comments!

 

[TeoAlfa]

And here are the CCT and TTT diagrams of the material as a reference. Again i am not totally sure about the material but i think it's something close if not 4130. Thanks!

 

[mrfailure]

I'd be looking at a more basic question: Was preheating performed BEFORE welding? You would certainly get cracking in the HAZ that could extend into the weld if this basic step was not performed for welding of 4130 without preheat.

 

[TeoAlfa]

Yes i forgot to mention that the areas to be welded were preheated at 150deg Celsius prior welding. Thanks!

 

[racookpe1978]

Pre-heating to 150 C (300 F) is about right, but only if it were a good, even pre-heat uniformly applied across the weldment that didn't contaminate the fitted surfaces, and if welding were begun immediately after pre-heat temperature is met and held through the weldment. Make sense?

A torch flame rapidly passed over the surfaces, blackening and baking in carbon soot across the inside of the joint for a few seconds would do the opposite, making cracking more likely.

Galvanizing a weldment at 450 C (842 F) may or may not affect it: certainly, if a thick, complex weldment were quickly lowered into a bath of liquid metal at 850 degrees from a cold pallet on a cold floor, the weldment would suffer thermal stretching and strain. WOuld it crack? Maybe, maybe not. Test a cheap piece of similar size and weight at the same temperatures BEFORE you try the galvanizing again.

A thinner weldment with "friendly" easy bends and twists will not crack: That why you see the old cast iron wheels and pulleys and shafts use those "ornate" curved legs and struts. Cooling a thick hub, straight struts, and a thick, heavy outer rim will stress the pulley as those thinner straight legs cool down so much faster than the hub and rim.

 

[TeoAlfa]

Another structure with not special steel but only S355 was also galvanised but no crack at all!
Does this mean that the crack has been caused by the stresses applied (during galvanising) to the already brittle (even partly) martensitic structure formed after welding?

 

[Cablini]

A joint between 4130 and S355 is very unlikely to crack (or form any martensite) based on the CE of each material. Add to that the fact you're using a 150C preheat and cracking from welding is almost impossible even without the use of a weld procedure.

450C isn't high enough to create martensite, you must austenitise first to be able to form martensite upon fast quenching in these materials.

It sounds to me as though the hot-dip galvanisation has caused hydrogen cracking. Hot HCl used in the galvanising process causes a corrosion reaction at the surface, producing hydrogen which diffuses into the steel, causing a significant amount of embrittlement. Any residual stresses from the welding or initial forging may then have caused the structure to crack.

Can you tell me what orientation the cracks are with respect to the rest of the structure and the welds?

The solutions i'd suggest would be:

Perform a PWHT to relieve all residual stresses
Don't use hot dip galvanisation (its just asking for trouble)

 

[metengr]

There has been liquid metal embrittlement cracking of structural steel with no PWHT as a result of galavanizing. You can search the internet to find several reports on this subject;

http://www.aws.org/itrends/2004/07/it0704-29.pdf

http://www.aimme.es/archivosbd/observatorio_oportunidades/failure_mechanisms_during_galvanizing.pdf

 

[OGMetEngr]

Everyone has pretty much covered the likely causes. The acid pickling prior to galvanizing introduces hydrogen, and any residual stress along with martensite formed by welding can be enough to begin cracking at that point. The 850F galvanizing bath will bake out any remaining hydrogen, but the damage may already be done. At that point liquid metal embrittlement is a possibility, especially if any microcracking or fissuring occurred during welding or pickling.

4130 can definitely crack if welded with no PWHT. It's CE is on the upper end of what's field-weldable, but with the relatively high level of carbon I wouldn't recommend it. I'd be willing to bet most if not all cracks were in the 4130 part.

 

[Tmoose]

welding process ?
welding parameters ?

 

[weldstan]

TeoAlpha.
You give no indication of the thicknesses of the parts. The thickness has a great deal to do with martensitic formation and hardness thereof in the HAZ of the 4130 alloy steel, and joint restraint / residual stress, especially at the toe of the welds.

 

[Guest102023]

metengr has identified the most likely problem; actually I was going to post one of the same documents.

Its an elusive problem that can drive you quite mad trying to eliminate.
Grind off flame-cut surfaces, have no sharp stress raisers, and check the silicon content of the base metal.


"If you don't have time to do the job right the first time, when are you going to find time to repair it?"