Fabricating I-Beam

[lancer360]

I have a supplier in Singapore that is having a hard time finding various size wide flange beams in low temp rated material. He has come back asking to fabricate the beams instead out of ABS EH36 material. This plate material is already approved by us and is used for many of the plate components of the job and he already has the necessary weld procedure also.

I know I have seen beams fabricated in the past, but I have never been directly involved. Is there anything special I need to look out for or special requirements I should impose? I was figuring that a full penetration weld with the necessary UT inspection would be acceptable, but figured I would pose the question hear and see if anyone had some guidance.

Thanks!

 

[msquared48]

If this material is comparable to standard A36 material, I don't see that the restrictions would be different than for any steel plate girder built here in the states.

Mike McCann
MMC Engineering

 

[SlideRuleEra]

You may want to reference the 2005 AISC specification, see this question & answer in "Modern Steel Construction" magazine:

http://www.modernsteel.com/steelinterchange_details.php?id=638

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[JStephen]

Ideally, you'd design the beams that way in the first place, rather than replacing wide flange beams. The web and flange thickness of beams don't match up to standard plate thicknesses, and the properties and design criteria are figured somewhat differently. You also have some differences in edge spacing at bolts. I don't know that the welds would have to be full-penetration.

 

[Lion06]

I agree with JStephen that the welds don't necessarily have to be full-length, full pen welds. I think that might actually be detrimental as you are throwing so much heat into relatively thin plates, you can get warping and a lot of residual stresses. It might be better to stitch weld them.

 

[kelowna]

If the fabricated section has comparable thicknesses and dimensions, I do not see a problem. In fact it should be better. Tolerances for rolled sections are worse than for fabricated sections.

You might have to check the bending calculations. Depending on the code, to calculate the bending capacity you use a parameter which has different values if the section is rolled or welded (sorry but I do not remember the specifics and I do not have the code handy.

Absolutely no full penetration welds. It is not an easy weld (on 'T'), the heat distorts the flange and it is not required to transfer the shear stress. The calculation for the shear transmission from the web to the flange is quite simple and you will see that very little weld is required.

 

[lancer360]

Unfortunately, I'm on a boat in St. Johns, Newfoundland on a different job and don't have access to any of my reference books or any of my calculations regarding the loading on the beam in question. I have to tell the supplier they can either keep trying to source the original W16x67 or wait until I'm done with the job and get back to the office and my files. Either way, I will research this some more as I have a feeling this is becoming the norm. It is getting very hard to find cold temperature wide flange beams and everything we design nowadays is designed to DNV requirements at -20C.

Thanks for the pointers.

 

[hokie66]

Good luck with the low temperature rated steel, a topic about which I know nothing. But one of the recommendations above struck my eye, about stitch welding. Since you are a marine/ocean engineer, that probably would be asking for corrosion problems.

 

[Qshake]

Also in addition to the comment on stich welding and potential use in marine engineering...make sure you have no fatigue related issues.

I have made I-shapes from very large plates when rolled shapes were out of the question but nothing as small as W16 type plates. This will force the fabricator to pay attention in the welding process to meet tolerances.

Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[EngineerTex]

I agree with JStephen and StructuralEIT about the stitch welds, but pay close attention to the comments by Qshake and hokie66.

In the past, in order to avoid the corrosion problem for built-up sections in marine environments, I have done the following:

1) Design the stitch welds as required -- for example, a fillet weld about half of the thickness of the web, and welded from both sides, to completely seal on both sides of the weld.

2) Open a gap between the web and the flange. If, for example, it's an 1/8th fillet of four inches long, with six inch spacing between each weld, cut a gap along the top and bottom edges of the web plate between the weld locations prior to fit-up. I usually put a generous radius of twice the thickness of the web to avoid sharp corners. (grind inside corners smooth)

3) Finally, weld along every part where the plates touch, being certain to weld closed each end of the stitch weld.

The description in number 2 may not have made sense, but what I'm describing is similar to how the web looks at the coped end when bolting (or welding) to a column. You just do that 20 or 30 times along the length of the web at top and bottom.

-T

Engineering is not the science behind building things. It is the science behind not building things.

 

[HgTX]

Between full-pen welds and stitch welds there's the much easier option of full-length fillet welds. Need to watch for flange cup or tilt, but this can be managed.

Hg

Eng-Tips policies: faq731-376

 

[GRoberts]

I agree with HgTX. Usually, the most economical weld to make is the smallest weld your can make (single sided if you can live with flange tilt, and double sided if you cant) in one pass. Typically the fillet welds don't need to be that big. If (based on strength calculations) the continuous fillet weld size gets too small to make, then you should look at intermitant fillet welds. Don't make the intermittant welds so short though, that the size increases enough that you can't make them in one pass.

 

[Dinosaur]

I don't know how long your beam needs to be. I doesn't sound like it would need to be very long. However, I would like to add a few of my own thoughts. First, make them submit a shop drawing for the new proposal for review. Second, make sure the flange and web are aligned so the 'direction of rolling' is parallel to the axis of the member. Third, use continuous fillet welds both sides to attach the flange to the web. In my work, bridge girders, fabricators like to work with thicker webs than the minimums required to avoid web buckling. If your piece is short, maybe that is not as important. I would recommend CJP welds if they splice web or flange plates together to make up the full piece length. Except for this I would avoid CJP welds. Good Luck.

 

[MotorVib]

Question why only -20C? I have work on projects in Canada and -40C is the low.

"In this house, we obey the laws of thermodynamics." Homer Simpson

 

[lancer360]

We only use -20C because our equipment is used offshore and it just doesn't get as cold offshore as it does onshore. Majority of the systems are in the Gulf of Mexico, offshore West Africa, Indian Ocean/Southeast Asia so they will never see cold weather. However, my client does operate some systems in the North Atlantic (Grand Banks) and the North Sea. Since they never know where a system will be in 5 years time since the rigs move all over the world, they design and build everything to DNV (Det Norske Veritas) with a design temp of -20C. They have been burned too many times in the past with a system that was not cold weather and the rig gets a contract in the North Sea and they have to rip out the entire system and replace it, or worse yet loose the contract to a competitor if we don't have a DNV cold temp rated system available.

Dinosaur, thanks for the tips on the construction. The longest continuous piece we need is 44 ft. so it will most likely have to be spliced. We have spliced the beam in the past and we have a particular area that is very lightly stressed that is convenient for the splice.