What type of steel is this? 4

[cookyb]

I've got a building built in 1993 that was built using (reportedly) recycled bridge steel (don't know when the bridge was built). We are doing a remodel and are having some trouble welding to the existing wide-flange beams (cracks have been visibly detected at all of the welds so far). I had them sample a couple of the flanges to have a chemical analysis done so we could try to determine exactly what type of steel we have and develop and weld procedure. It's a 3-story building and samples were taken from the bottom flanges of random beams at each level. After the analysis, the testing firm reported that there were actually two types of steel being used, 1018 and 1020 steel. Attached are the test results.

As you can see, there are a couple of other chemicals floating around in the steel sample as well. The testing firm, who is also acting as the CWI, tried a test weld by preheating the sample to 225 degrees F at approx 3" from the weld. During their bend test after welding, the sample cracked and broke in the preheated portion of the sample, not at the weld, and at not much of a bend angle either (see pic).

Are these results indicative of 1018 or 1020 steel? Has it been carburized and that's what is causing the welding difficulty and extra chemicals? My main issue is developing a weld procedure for the old steel. So far the testing firm has tried preheating and welding with a flux-core wire feed.

Any ideas?

 

[redpicker]

The "other chemicals floating around" are not of concern. These are often termed "residual elements" and the levels observed are normal.

The Silicon content is interesting. It is normally in the 0.15-0.35% range, unless purposely kept low. It is usually kept low for weldability reasons, so this is not the issue, but it is interesting.

What is not reported in the levels of Vanadium (V), Titanium (Ti), Boron (B), or Columbium/Niobium (Cb/Nb--these are just different names for the same element). These are often used for added strength and their presence in the 0.03% level or higher could cause trouble (0.0005 for Boron). Just taking a wild guess, this is what you are dealing with.

rp

 

[jwhit]

I agree with Redpicker about the residuals, it is the one's that aren't reported that are of concern. I don't see rock candy in the fractures, but would also take a good look at your aluminum content in case you are dealing with some aluminum nitride. Also considering the bridge material probably goes back a long time and melting/pouring methods weren't what they are today, you might consider residual gases as a cause for poor weldability.

 

[cookyb]

Thanks for your help, guys.

Redpicker, would Vanadium, Titanium etc normally be tested for in a chemical analysis or do you have to specifically test for each one?

Jwhit, I thought preheating the weld was in part to drive off any residual gases. Is this not the case? Also, I'm not familiar with aluminum nitride; why would it be present?

I should also tell you that the Rockwell B hardness tests for the two materials came back ranging from 70-83. Do you think we are still dealing with 1018 or 1020 steel? I've never heard of it being used for w-flange beams before.

 

[weldstan]

You have a semi-killed, A-36 steel with large Fe/Mn sulfide type stringer inclusions, some of which are visible near the surface. Do take an Al analysis plus V and Nb.

 

[metengr]

The sulfur content could indicate you have inclusions resulting in laminations that are interfering with welding. I have seen this before in older steels.

 

[weldstan]

When I was a steel mill metallurgist from '68 to '72, thinner thickness A-36 was commonly made semi-killed (.010 Si max.). We also downgraded materials into A-36 when ladle reactions (exothermic) occurred. A-36 was truly the "bottom of the barrel"; the only grade lower was "fence post". These materials would have multi-layers of sulfide stringer inclusions.

 

[cookyb]

I had assumed it was A36 during design and was pretty surprised when the test firm said it wasn't. Sounds like it still may be, but just on the crummy side, huh? Nice that we have slightly better than "fence post" steel holding this thing up.

As far as a plan of attack for getting a good weld on this, what would you recommend? Do we need the additional testing (Al, V and Nb) to figure this out?

I really appreciate your help, guys.

 

[metengr]

If you can remove a small sample for metallographic examination, this would confirm if you have stacked laminations. Welding will require more of a technique change to butter the surface with overlapping stringer beads to build a base for welding.

 

[cookyb]

The most critical welds I'm dealing with are 3/8" flare bevel welds to attach a 25' long HSS5x5 to the bottom of the bottom flange of the beams in several locations. The CWI has said that the stick welds of misc steel to these beams have looked acceptable, but these are all VI. They tried a line feed welder for the tubes and started running into trouble with cracks forming immediately. The CWI was skeptical of stick welding the tubes because they were worried of cracking of the welds due to starting and stopping while changing out sticks.

 

[weldstan]

For best strength, you need to butter the flange with E7018 low hydrogen electrodes or other low hydrogen welding process with 2 layers prior to making the strength weld to the buttered layer.

 

[racookpe1978]

You said in the first post that this was "re-cycled" bridge steel.

two things: I understand this means the bridge steel was taken down and rebuilt, NOT re-melted into new shapes, right?

If so, is there surface contamination and residuals on the weld areas (old salt, paint, oil/grease/tire rubber/atmospheric "gunk" maybe) that is fouling up your surface before and during welding? If you are grinding clean the weld areas, how "clean" are they, and how deep is the grinding "on average" and "in actual practice"? Just "shiny" may leave lost chemicals on the surface, smearing them into the metal pores.

 

[cookyb]

racookpe, yes. The bridge was disassembled and then the pieces used as-is. It's really an odd way to put a building together. Columns are not continuous, they stack on top of the girders. Most of the girders are 3 beams side by side and stitch welded together...weird. And yes, there is tons of welding that's already been done, nobody seems to know how though.

The steel all had lead paint on it that was removed by sand/water blasting. I'm not sure that they have done any grinding to try and clean the surface. I'll check and see.

 

[mrfailure]

This steel had been in service in a structure and may well have inherent residual stress that caused the cracking. The stresses may have been from installation as well as developed over time as the bridge was settling. I am assumming there were no preexisting cracks in the bend test sample (that would be a different discussion). I would take the steel to a heat treater to at least stress relieve if not anneal it before trying to use it in a repurposed application.

Of historical note: A36 was first published in 1960 and replaced ASTM A7, which was structural steel used for bridges and building. A7 may be your applicable spec depending on bridge age.

 

[cookyb]

Thanks, mrfailure, but I don't think it's A7. The chemical analysis doesn't match up. A7 looks like it has 10 times the amount of carbon and 200 times the chromium of my sample.

This building was built in 1993, so there won't be a possibility of any heat treating. No existing cracks in the steel have been found. Unfortunately, we have no idea when the steel was originally put into use.

 

[Guest102023]

Dirty secrets of the construction industry.
Let me know where that building is so I can put it on my list of structures never to enter.

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

 

[weldstan]

During the late '60s and early '70s Interstate 5 was being built in CA and a number of steel bridges came a tumbling down during or just after construction, usually at temperatures around 40F. We always breathed a major sigh of relief when we determined that the steel was not ours.

 

[racookpe1978]

Most of the girders are 3 beams side by side and stitch welded together...weird. And yes, there is tons of welding that's already been done, nobody seems to know how though.

The steel all had lead paint on it that was removed by sand/water blasting. I'm not sure that they have done any grinding to try and clean the surface. I'll check and see.

Hmmmn.

If a significant lead layer were still present after waterblasting - because I know shotblasting if done poorly will only drive paint articles deeper into the pores and scratches in old steel - would that lead screw up the welds towards the breakage you see? .. I've never gotten a satisfactory weld (GTAW especially) on old painted surfaces until they were ground "bright clean" (rough is OK, but they have to be clear metal. Whenever possible, even after paint removal, I specify a qualified SMAW (stick) weld because the greater heat burns through better.

Obviously, not all of the welds broke off. Is there any pattern to which ones broke and which ones were good... so far good, that is?

 

[cookyb]

So far, other than the 3 test samples I sent, only the wire feed welds have cracked. The stick welds (mostly puddle welds of deck and welding angles to the existing girders) have been fine based on a visual inspection. I was at the site yesterday and we decided to try out welding the tubes via stick welds and then testing them. I'll recommend that they grind the welding surface to a clear metal condition as well. Thanks for your ideas, everyone. I'll keep you posted.

 

[weldstan]

"Only the wire feed welds have cracked." Solid wire (short circuiting transfer?) or flux cored wire (self shielded or gas shielded)? AWS classification?