So, you walk into a job on Monday morning...and find this on the floor (See attachments). 

You look up and in the side of the member you can see all the way through via a 3/16" wide crack.

I'll add the proverbial other shoe that will drop later.  I would like some comments based upon this part to see what your thoughts are.

He Is In Control, Have a Great Day,  Brent

That's all you have for us?

The lower portion that is brown, it that corrosion?

The blue region appears to have occurred at an elevated temperature, somewhere around 575 to 600 degrees F.

How about letting us know about the thickness, material specification, grade, etc.

Reminds me of a photograph I received as an attachment from an engineer I went to school with. The photograph was of an end plate moment connection that failed during erection. The e-mail went something like this; "Hello Al; It's been a while. Please see the attached photograph."

I wrote back: "It cracked."

Are you related by any chance? Nice photographs by the way. They show some nice detail.

Best regards - Al

Okay, Okay, I'll give you as much as I can but pictures are out because this prima donna shop thinks all of it's structural work as well as their welders and other equipment is top secret rocket science grade stuff.  They allow no pictures to be taken in their facility.  I'll not use any names and to their credit it is not something you can point a finger at and say they purposefully messed up.  Let me explain:

Box columns, 3" opposing sides and 2 or 2 1/2" on the other opposing sides.  A572 grade 50 plate steel used.  All material verified. 

100's made and then a change, the 3" sides cut to allow for a 6" plate to pass through and get welded with CJP's.  Same grade, about 36" sq and 6" thick.  Welded to the 3" side plate.  Come in on Monday morning to a surprise. 

D1.1 minimum preheat.  FCAW with 3/32" E70C Lincoln electrode. 

Now, I know how I would have handled this.  How about others?  Both the welding and the explanation for what happened.

He Is In Control, Have a Great Day,  Brent

Okay,

Kent, Al and C.Tacker are all thinking along the same lines I am especially since I know this shop and how things are normally done.  And, not being a metallurgist I have been going back through Linnert's 'Welding Metallurgy' which is very good but takes a novice a bit of chewing to get into. 

Oh, but I like Martin's comment best. 

I'll explain more tonight when I get to the job. Gotta go for now.  Still moving, actually, mostly remodeling.  That's what happens when I leave my wife and son home alone to start getting things ready.

He Is In Control, Have a Great Day,  Brent

If I can, I will accomplish this post without interruption,    

So, if you are like me, the metallurgical aspect of welding has interested me for quite some time.  Along the way in the past 40 years I have learned a lot and ran across a lot of misinformation, some of which, sadly, I believed for far too long. 

Coming across some books on metallurgy, including getting a copy of Linnert's book when I joined AWS many years ago, then, going to the CWI Seminar and getting my CWI, much of the misinformation has been clarified, corrected, and replaced with truth and has been of much help to me in the past few years. 

I still come across many things that make me ask, 'What!?  Why? How?'  There are many things that do it but there are some things I just can't let go of.  To pursue this one is of particular interest but also difficult because of the restrictions on photographs. 

So, after personnel was shifted around, one of the shops I have been busy with over the years had this occurrence back in February, if I remember correctly.  Time period of importance to my train of thought on this.  'Why?'  WEATHER.  TEMPERATURES. 

I have been able to follow this from a distance and a couple of times be called down to stand in for one of the other inspectors on the job and saw some things first hand.  I still am going to try to refrain from making statements that would get me or anyone else here in trouble.  I just want to know what happened. 

So, FYI, I am referencing 'Welding Metallurgy' Volume 1 chapter 3- The Properties of Metals.  It has been very interesting as I am understanding more of it this time through than I did previously.  Especially when I came to the parts on Brittle Fracture and Ductile Fracture.  Then he added temperature factors and when he started discussing Elevated Temperature Service and High-temperature Service Some things really started to click.  And, right when I got sent back to the shop where this problem was most obvious.  Second note of importance: there were several cracks involving several members of this job, not just one or two. 

For those who may not have Linnert's book for reference (which is pages 156-180 specifically though the entire chapter would be good for a proper background), the third item of importance based upon wording on page 165 is the coloring of the material at the fracture.  To quote, "The surface of a high temperature fracture in iron or steel that extends to the component surface and allows air to enter the crack is usually brown or blue because of the formation of temper colors." 

Now, according to the text just prior to that quote concerning a description of High-temperature Service the loss of strength happens rapidly as the temperature exceeds about 750°F.  The explanations around this topic explain how it can undergo changes beginning at 500°F.  Changes to toughness and other properties.  Sound familiar?  How about the 550°F limit to inter-pass temps incurred by D1.8? 

Anyway,  my mind starts going through rapid spinning and smoke exhausting (a normal condition according to my wife and kids).

See, knowing how most welders and the shops they work at do things, I am forming some ideas that I need to run past those far more intelligent and educated than myself, especially in the arena of metallurgy. 

D1.1 establishes a MINIMUM pre-heat for the materials used in buildings.  For the A572-50 involved according to Table 3.2 in D1.1-2010 our materials in question only need to be heated to 225°F.  Both the 3" and the 6".  Now, how many times have you seen this done properly?  Especially without the TPI standing right there the entire time?  Then, think about the item we are dealing with, a box column with 2- 6" solid slabs running through it (really wish I could use a pic, think I will draw this out later and scan then attach it. Not much of an artist but I bet I can make it more clear that way). 

So, next item of importance, the part configuration and the stresses built up during welding.  Remember, we will also have the four corners of the box column welded by SAW with CJP's as well as these plates running through the box.  Add to that the issue of pre-heat.  Heating this member in a small area around where the weld will be made and expecting it to hold that heat is really an effort in futility.  This will leach off, can we say, convection, radiation, and how many other proper terms for the way the heat is going to be drawn off, our preheat so fast there is no way most shops would have considered this at least on the first one.  Plus, they already had their approved welding procedures and shop drawings with no additional restrictions or requirements in the Contract Documents.  Not to mention that no one thought about PWHT for this material grade even with the thickness and stresses involved. 

One last item, maybe.  The weld access holes in the sides of the box column.  WAY too tight of a radius considering thickness and stresses.  And, they did a good job of cutting them and cleaning them to bright surface, BUT, with no one experienced in metallurgy and familiar with some of the items mentioned in the Fracture Mechanics section of chapter 3 of Linnert's book, there were way too many stress raisers (notches, changes in thickness, even 'cracks').  Oh, they were within the tolerances of Section 5.15.4.3 for roughness/smoothness.  And even with extra given for D1.8 and going to Sample #4 on the C4.1 guide.  And then ground smooth and shiny.  But, with these combinations of stresses all built up any change allows for a stress raiser that leads to a Stress Multiaxiality (big words that I don't fully understand but it is all making more sense as I see it in good application) in this case being a triaxial stress which is so because of the irregular surface of the weld access hole radius not being perfectly even and smooth (better to have been drilled and then cut the straight sections with a torch (which also may have started the problem is proper preheat was not done prior to cutting, check that one out in the codes). 

And, I lied, this will hopefully be the last item of interest but, it was also hinted at in the beginning.  The WEATHER.  So, without proper preheat, with triaxial stresses, and with all the other factors of heating, cooling, heating, cooling, and apparent from our reference material heat in excess of 750°F (which I have seen shops do when putting multiple welders on a section too close together and with this big wire, 3/32) the part was taken above the critical area and then at the end of shift, eveyone left with no blankets or other source of heat control used on a weekend that even for the AZ valley region was 'cool'.  The members cooled way too fast and uncontrolled. 

Part of this will make more sense once I explain how the engineers regulated the procedure when alerted to the problem.  Multiple heaters with blankets and other resources to take the entire member to 300-400°F (prime= 350°F).  Do all work on member relating to these joints with the 6" to 3" plates.  Take on up to 500°F and hold.  Start the downward heat flow until down to ambient.  Shifts went to 24/7 and stayed that way for about 3 months while this was taken care of.  Once a member started preheat it stayed in the loop and wasn't allowed to get below 300°F until all work was finished and the 500°F hold point had been reached. 

Then, they opened up the radius on the access holes and ground them really smooth and even.  Used MT, UT, and PT to check areas over and over again. 

So, here is my reason for posting.  Based on the above run down of all being done, does my 'theory' look valid?  That being, my interpretation of Linnert's book and the brittle fracture from high temperature service and the application I have made to this current problem?  Or am I way off base?

I am not presenting this to anyone nor making any legal standing, I am just curious and want to know what happened and why it happened? 

He Is In Control, Have a Great Day,  Brent

My hesitation is a detailed response was well founded. Way too many unknowns to draw any conclusions.

I think your point on triaxial stresses may be moving in the right direction. The stresses are the result of the residual stresses developed during welding. They can be added as vectors, so intersecting welds can produce resultants that exceed the tensile strength of the base metal.

Considering the thickness of the member’s involved, low ductility, low notch toughness, and brittle fracture are high on the list for probable contributing factors.

The preheats listed in table 3.2 are the minimum preheat temperature requirements and do not take into consideration the degree of restraint. A more conservative approach would to use Annex I. It includes the degree of hydrogen control, chemistry, and degree of restraint. The recommended preheats are invariably higher than table 3.2 for highly restrained joints. In this case, I would predict the contractor could meet the H2 criteria, but not the H1 conditions.

We still do not know the entire story because we do not know the connection details and the degree of restrain. However, given that the members are thick, they would meet the criteria for high restraint.

Preheat is always an issue because, as you point out, contractors typically do not preheat the material in the through thickness direction or laterally from the groove edge for the distance required.

On the subject of residual stresses, the temperature required to develop stresses in excess of the yield strength is relatively low, on the order of 220 deg. F. Taking that into consideration, add ambient temperature, any base metal heated to 300 degrees F can be assumed to have a residual stress on par with the yield point of the base metal. It is easy to exceed the tensile strength of the base metal when there are intersecting welds. Alloy steels can also experience “blue brittleness” at the temperature ranges that can be attained when multipass welding techniques are employed without regard to interpass temperature.

There are a number of issues that need to be considered, but without knowing all the facts, jumping to a conclusion is for the simple minded.

I am still unsure about one fact, did these failures occur during fabrication or in-service? You didn't clarify the issue of the discoloration. The brown, was it temper color or corrosion?

I do like your approach and I also use George's books all the time.

Best regards - Al

To Al especially but also including all the rest who have responded.

I know this has been a little off the wall but my curiosity would not let me let go of it.  I needed to be able to take this and make it as much of a learning opportunity as I could and I hope others may pick up a few points from it as well. 

Al, your answer is very good.  Tyrone, yours too.  Others are right in there with their views but I know my info was limited.  Rather odd circumstances. 

If something comes up and I feel free in the near future to add more info and especially pics I will. At least the drawings I mentioned may be of some help.  I'll see what I can do. 

Meantime, the additional pointers on the application of the things I read in 'Welding Metallurgy' has been enlightening and is much appreciated.  You also gave me some ideas that I will try to get answers to in order to fuel this a little more and get more specific. 

Thanks again.

Brent

Any updates?

Al