Welding Wide Flange Splices 4

[sedonas]

When you weld two wide flange together by cutting them square... would it make sense to further strengthen the web by putting cover plates over the web? Or over the flange?The purpose of flange is for tension and compression.. the web is supposed to be only for shear? Which to focus more in welding.. But if you weld cover plates on the flange.. won't it weaken the flange by having many welded regions.. how do you weld your splices (that involves no bolted connection)?

 

[sedonas]

Is your weldor trying to do each weld in one pass?
What is the welding machine they are using? What are the settings?

I don't know the setting. They probably just estimate spark intensity. The old welders won't be back because in our place.. once welders got the money.. they never got back because the money is only good for a week for their family then they are broke again. Yes they are that poor. This is true for 99% of welders in my country.

Well. For the next welder who will repair the splice. I have to verify every step he made. For a 10mm flange thickness.. what is the usually the size of the bevel?

The old welders spent 1 day trying to add welds to half inch of gap for rafters than frame into the wide flange body.. imagine connecting web to web perpendicularly.. good it is on top of column so not hanging.

How big a gap before you can fill it with welds? They spent one hour just trying to build welds for each half inch (12.5mm) gap. They probably ignored the flange thinking the main support of wide flange are in webs.. so they focus on webs mostly.

Thank you all for the urls and videos. Will study them so make sure the repair team will do it right. Again. What is the usual bevel size for full penetration weld.. any table for given depth or thickness and bevel size?

The flux etc needs to be cleaned off for proper inspection.

Here is Lincoln Electric's basic video about visual weld inspection. It is aimed at Boy Scouts getting their welding merit badges.

https://www.youtube.com/watch?v=Ncguc7THEUY

I think the "welder" needs at least to spend a few days running beads flat and vertical and making pad welds on 10 mm material to achieve at least basic proficiency. Even then, I don't think he should be allowed to work on your job again.

I think the "expert welder" is going have to do a LOT of grinding or arc gouging to fix those welds. And probably cut out and replace some sections entirely.

Pages 32 thru 33 here discuss a minor case of one of the obvious deficiencies of those welds you are having to deal with.

http://supplier.huntingtoningalls.com/sourcing/doc...

Note the corrective action of cleaning the weld area.
The lack of cleanliness is NOT causing most of the problems however.

I envision the beam welding requiring be done in 3 different positions. Flat, vertical, and even overhead.

Here is a video showing some very good "vertical up" welding in process, and proper surface and edge preparation for 3/8"/~10 mm plate. Full penetration welds generally require some beveling or other edge preparation and multiple passes. They are NOT done by blasting thru the material in one pass.

https://www.youtube.com/watch?v=1QxT7JUjs94

Overlook the extremely uniform weld surface appearance. There is some leeway there.
Can you see how the weld completely fills the space between the plates, and fuses completely with each plate?

here is another series, from Lincoln electric. OK, but they spend too much time recording far away.

https://www.youtube.com/watch?v=oCS9t2CB6Vg

They are all using 7018 electrodes. As I recall the flux removal is much easier than 6011.

But notice the uniform bead, slightly convex, fusing with each side of

 

[moltenmetal]

There are others here much more knowledgeable in structural welding than I am, so take their advice first- but seeing the pictures you've posted has me worried- you need to take care of this!

With two pieces of 10 mm thick steel butted tightly together without any additional field preparation, there is no way you can deposit a reliable full penetration weld. If you want metal from the welding electrode to flow into the joint, you need space for it to flow into- otherwise, you are trying to use a stick welder to do an "autogenous weld" (where the parent metal of both pieces is melted together and fused)- that just plain doesn't work. Just because you see "burn through" at the other side of the joint, does not mean you have deposited a sound, full penetration weld at every point in that seam.

It is my opinion that the right preparation for this weld would be to bevel both pieces on an angle with a grinder from the outside, leaving a vee shaped groove all the way along the joint, leaving a small root gap between the two pieces at the point where the edges meet. The other option would be to try to do a partial penetration weld from both sides, grinding out the root of the 2nd weld after the 1st is completed to get a clean surface on which to carry out the 2nd weld. However, to be sure you did this properly, you would still need to do sufficient grinding so that you ensured you got a full penetration weld at the point you can only reach from the outside, i.e. the point where the web joins the flange. The vee groove and completion of the weld entirely from one side is just easier.

If the welder is worried about burning through at the root, leaving a mess, then a backer strip can be used. However, the presence of the backer strip makes it difficult to inspect the root of the weld. A good welder would not need one, but depositing the weld on the lower flange in place (i.e. welding 100% overhead)takes considerable skill and practice.

Any of these processes are beyond the skills of the person who attempted to weld this mess in the first place. You need a real welder, trained and qualified, to do joints like this.

 

[sedonas]

There are others here much more knowledgeable in structural welding than I am, so take their advice first- but seeing the pictures you've posted has me worried- you need to take care of this!

With two pieces of 10 mm thick steel butted tightly together without any additional field preparation, there is no way you can deposit a reliable full penetration weld. If you want metal from the welding electrode to flow into the joint, you need space for it to flow into- otherwise, you are trying to use a stick welder to do an "autogenous weld" (where the parent metal of both pieces is melted together and fused)- that just plain doesn't work. Just because you see "burn through" at the other side of the joint, does not mean you have deposited a sound, full penetration weld at every point in that seam.

The above was exactly what the welders want me to believe.. that my using the powerful 6011 electrode.. they can fuse the 10mm thick steel even without any gap for the metal from the welding electrode to flow to..

They said the welding arc melts the entire 10mm from one side and the filler filled it up even if no gap..

But we not into welding thought it was logical the welding can melt thru 10mm and fuse them.. but theoretically why doesn't this occur? Let's consider a welding electrode without filler.. if you touch the 2 pieces boundary without gap in them.. if the welding is very powerful. won't it melt the 10mm thick metal and somehow fuse them? Is it a limit of welding machine that makes this not possible... maybe there are very high tech welding machine that just can do this? What principle avoids it if not at all possible at all?

If I can't find local expert to repair it. I may bring in international welder to fix it.. but there is only 4 splices so the cost may not be friendly.. Hmm.. any international welders here coming in for vacation? Maybe can broke some good discounted deal...

 

[gtaw]

There have been many valid points made in this thread. For structural application, AWS D1.1 has been the main stay in the US since the early days of steel framed construction with welded connection. The standard has undergone many revisions and it has incorporated many "lessons learned" from failures, both large and small. It is the go to document when welding steel framed buildings. I urge our friend Sedonas to purchase a copy and follow the requirements.

One issue that wasn't mentioned was the base metal being used. If it is ASTM A992 it should be welded using a low hydrogen electrode, i.e., E7018, not E6011. If it is the equivalent to ASTM A36, it can be welded using E6010 if the thickness is 18 mm (3/4 inch) or less. By the way, E7018 must be kept in unopened hermetically sealed cans until they are needed. They must be stored in electrically heated electrode ovens to ensure the moisture levels are kept low enough to meet low hydrogen requirements, i.e., less than 16 ml of diffusible hydrogen per 100 g of weld deposit.

The attached sketch depicts a typical groove detail that is applicable to the beam splice and the subject of this thread.

If skilled welders cannot be found, bolted connections work. Bolted splices are used in bridge construction all over the world.

Best regards - Al

 

[bb29510]

dont forget even a A36 if its a grade 50, needs a low hydrogen rod.

master ICC inspector, AWS CWI

 

[gtaw]

AWS D1.1 doesn't list ASTM A36 grade 50 as a prequalified base metal. I believe it is the intent that ASTM A992 fulfill that need. As such, all A992 requires low hydrogen electrodes.

Best regards - Al

 

[moltenmetal]

gtaw: the 1/4" root gap looks a little large for 10mm (3/8") thick material, but I suppose that's because there is a backer bar being used. Otherwise that's the prep I had in mind.

The AWS prep assumes that the weld of the lower flange will be done from the upper (web) side, which is much more likely to give a good result especially with a welder who is not very experienced. The only trick is to ensure that there is full penetration at the web to flange junction in this location, which is why they have ground away material at these locations- to give access for the weld metal to be deposited.

As long as the welding is done in the correct order, and the slag is cleaned between passes, this is the right way to do a splice joint in member like this by welding.

I agree with GTAW: if you can't get proper qualified welders, you can find bolted splice joint designs that can be used instead. They are far more likely to give a good result than a joint which is improperly welded.

I'd suggest to the OP that he consider getting a copy of AWS D1.1.

 

[sedonas]

I told the contractor who is a civil engineer to read this very thread. He still tells me he uses 6011 instead of 6013 precisely because the 6011 can penetrate the 10mm flange and fuse the inside of it even if there is no gap. Can someone explains why this can't work? Because by logic when you increase the temperature of the region to thousands of Fahrenheit.. it should become molten.. so he reasons the 10mm thick flange becomes molten and all parts fuses using the 6011 even if there is no gap.

 

[kingnero]

do the test.
have him weld a piece, cut it through, etch (google this) and see for yourself.

I strongly suggest, as said above, to get the relevant codes (either AWS or european), as most of what is said in this thread is basic knowledge and is imposed by codes.

 

[gtaw]

I agree, there is no better proof than welding a sample to see the results. There is no way the welder is going to penetrate 10 mm of metal using SMAW without some type of groove preparation.

While E6013 does not have the "digging" action of E6010, the 6010 isn't capable of penetrating that thickness without a proper bevel.

However, this isn't the first time where I've see people make assumptions based on nothing more than wishful thinking.



Best regards - Al

 

[sedonas]

He said he used very strong current in the welding machine and combination with 6011 just can penetrate 10mm of flange and turns them into molten and fuse everything in the middle. I need to give him theoretical explanation why this is not possible. We don't have the technology to etch and he lives a bit far for this experiment.. unless I can convince him using scientific reasonings why 10mm can't be penetrated without gap even with maximum current and 6011. Can anyone explain why it can't do in theory? Is it a limit of the welding machine?

 

[moltenmetal]

Tell him to prove his proposed procedure by taking two pieces of 10 mm plate, butting them tightly together and doing a weld without backer strip (so the root (back side) of the weld can be examined). The specimen then needs to be sent for visual examination by a qualified welding inspector, radiography, root and face bends and tensile tests in accordance with AWS D1.1 or any other structural welding code.

It is extremely likely to fail both the radiograph (due to lack of fusion), and the root bend test. But if it doesn't, you write down the settings of the welding machine, the type of electrode, the base metal used etc. etc. and you now have a qualified procedure and a welder qualified to that procedure. You can qualify new welders to that procedure using a subset of the tests listed above, typically just the visual inspection plus root and face bend tests. Others can give you the list of the rules, or you can read them in the AWS D1.1 standard- again, this is not my area of expertise.

A joint with lack of fusion flaws can be very prone to cracking, because it has a crack initiation site built in- that's why it matters!

There are very good reasons that butt welds in structural steel members (and in pipe) are prepared the way they are. Autogenous welds are possible, but not with a stick welder! The ones I've seen done are done for much thinner and simpler sections, using automated equipment and very different welding methods (tungsten-inert gas (GTAW) as an example). If a stick welding procedure without root gap or bevelling gave adequate and reliable results for butt joints like this, nobody would bother with the grinding and other preparation. There is plenty of motivation to find better, quicker procedures which take less labour, and if they existed they would be used.

 

[kingnero]

He said he used very strong current in the welding machine and combination with 6011 just can penetrate 10mm of flange and turns them into molten and fuse everything in the middle. I need to give him theoretical explanation why this is not possible. We don't have the technology to etch and he lives a bit far for this experiment.. unless I can convince him using scientific reasonings why 10mm can't be penetrated without gap even with maximum current and 6011. Can anyone explain why it can't do in theory? Is it a limit of the welding machine?

there is no technology in etching, I do this at home with a grinder and acid that is in household products.
you don't need to give him a scientific reason, when this could already be solved by a test that takes 15 minutes - all in. And that will close all discussion.
That will yield more result than this thread, where you are given all kinds of good advice but where you choose not to listen.

and once more - 6011 is not to be used for structural welding. see above.

 

[sedonas]

Got your point, the safer would be to use bolts and cover plates. If the surface of any existing welds in the flange splice is grind with grinder to make the cover plate stick to it flat.. would the procedure makes the welded part weaker? I mean.. if you grind flat any standard beveled butt weld.. would it affect the weld strength? This is to avoid hoping the splice would become stronger by being both cover bolt plated and welded at same time (by ignoring the welding part as contributing any strength).

By the way.. the manufacturer of the wide flange says the material is A36 & SS400 (Special order: A992). What is SS400?? Does it require low hydrogen electrode.

In the US. A36 were replaced with A992. In our country. If it happens and total lack of low hydrogen electrode. Then our structures would even get weaker.

 

[gtaw]

Based on the welds in your photographs, whether you use E6010 or E7018 is a moot point, akin to comparing a 4 cylinder engine to an 8 cylinder engine when there is no gasoline to run either engine.

The shortage of qualified welders is a problem in all countries at this time. Go with bolted connections when there is a question of the availability of qualified welders.

If you still use welding as the option of choice, ensure the welders are qualified by making them pass a test before allowing them to weld on product.




Best regards - Al

 

[sedonas]

and once more - 6011 is not to be used for structural welding. see above.

Where is the "above".. I thought Ron said something along the line of "As for the electrode, an E6011 electrode is a penetrating electrode that can be used in all positions. It is a versatile, "all purpose" electrode. It provided deeper penetration; however, it also yield greater porosity and slag inclusions that can lead to nondestructive testing rejection. The welder must know how to use this and other appropriate electodes for the application."

What do you mean (and where is it said) the 6011 is not to be used for structural welding? What other substitute would we use then? Were you referring to low hydrogen 7018? But this is not available locally so all structures use the 6011.

I'm losing confidence in welding.. but there is another problem with cover plate bolted connections. Almost no one use it here to connect splice because they rely on welding.. so I may be the only one in the country who would attempt it. Many structural engineers don't know how to design it. For a w8x1.. usually how big should be the bolt? Since a bolt alone has 36 ksi.. maybe one bolt per side totally 4 bolts is sufficient in the splice? Any cover plates codes about it?

By the way, we don't have any welding code locally so liability is absolutely zero. The welders can't even afford a motorcycle to travel to work, they only commute.. so can't expect any payments for any liability claim.

 

[Wallache]

You lack basic knowledge of welding and bolting and yet you designed a steel structure that is currently under construction?

 

[gtaw]

I am not going to throw stones at anyone's glass house. I've seen some pretty ugly welds in my own back yard. Fortunately, I am in a position where I can insist the bad welds (or member) be removed and welded by someone with the require skills. The contractor may have to pay more to have it done correctly, but at least it does get corrected.

Best regards - Al

 

[sedonas]

The contractor may agree a repair on one of the splices. He plans to use acetylene torch to cut the flange welded region to create a bevel and reweld it. He doesn't have a grinder. He used entirely acetylene torch in all the cuts in the past job. What is your say about acetylene torch to make bevel or gap in the 10mm flange (as prep)? Is it ok?

 

[gtaw]

The oxy-acetylene torch is fine for removing the old weld and preparing the bevel, but the edges have to be ground to bright metal before rewelding the joint. The root opening (the gap between the two members) is going to be rather wide. Backing should be used.

Best regards - Al