Field welding to Exst member under load

[TheLinker]

I have an existing 36WF300 member (A7 steel) supporting a regenerator vessel and was asked to look into the existing shear stress at the copes. The web plate flexural stress is approx 30% over allowable due to the loss of flange steel at the cope location and a reinf bar has been proposed at the cope location to strengthen the section (see attached).

However the vessel load cannot be removed and shoring is not an option for the duration of the welding of the new reinf bar. Having referenced AISC first quarter 1988 Field Welding to Existing Steel Structures it indicates an AWS recommendation of limiting calculated stresses to 3ksi for unshored welding and that this is a conservative approach. Even if product is removed from the existing vessel I will not be able to get down to 3 ksi.

Does anyone have more information on allowable stresses in existing members during welding?

Any other ideas or insight on how to weld to steel that is at or near maximum stress or any other ideas on how to reinf this cope while under stress?

Any advice is appreciated.

Thanks
Roger

 

[ToadJones]

I guess the web would actually have to crack in order to mobilize your reinforcing plates then, right?

 

[BAretired]

The web would have to yield flexurally in order to mobilize the reinforcement plates. If it didn't yield, the web would continue to carry load just as is it has done for many years. But the safety factor against collapse would be improved drastically.

Why do you think the web would have to crack?

BA

 

[BAretired]

Change "drastically" to "dramatically".

BA

 

[TheLinker]

TJ,

I see what your getting at. My thinking is that the current web is overstressed (by 30%) which includes a reduction in thickness for corrosion allowance which has not been verified via testing, so the existing condition may be slightly overstressed but within the factor of safety of the orig design. The current support has been in service for 60 years plus and we are not increasing the loads on it, only reporting on the current state and recommending possable modifications.

I think that if the bars are added, the current condition is not changed, but if an increase in load does happen (large wind gust, etc..) and the web becomes overstressed to a point of failure the reinf would become active and act composite with the web when a larger deflection compared to when the rebar was added was experienced.

I am still mulling all this over and looking into BA's suggestion.

 

[MiketheEngineer]

A regenerator is often used in making gasoline. They typically weigh in at 500,000 to 1,000,000 lbs!! Yes - there are cranes that can lift them. And when they do - eveyone holds their breaths and crosses their fingers!! QUITE A SIGHT. We have shored a few of them while under going repairs/retrofits.

If this was set so long ago - why are you now worried about the beam capacity?? Has something changed??

If it isn't broken don't fix it!!! That said, you may have discovered a long exisitng and worrisome problem that does need attention.

Shoring posts do exist where you could probably get most of the load supported while making repairs. Also, there are hydraulic jacks that can be fine tuned using weight cells that can tell you exaxctly how much load you are lifting.

All of this is rather expensive - but it can be done.

 

[MaddEngineer]

TL,

Could you add a heavy W8 stub bracket welded to the top (or bottom) of the weak beam. Size it to transfer the required load to the web of supporting W36 beam, may have to be 24" long or more.

 

[paddingtongreen]

One of my first drawing jobs when I got on the board, was the riveted plate girders of the Catcracker structure for Fawley Refinery, oddly, they gave it to me because it was simpler to draw than the circular platforms and ladders on the towers.

For anyone not familiar with these, the two vessels were connected by 56" dia. pipes. The stems on the valves were over 20" long.

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[TheLinker]

Mike,

Nothing has happened other than reports of other similiar vessels of the same time period having issues/ cracking in the webs with large copes that may not been accounted for properly. You are correct about the estimated weights, we have estimated (on the high side hopefully) of about 1,500,000 lbs to 1,800,000 lbs for calc purposes so hopefully the actual stresses are lower then we determined. Build it bigger stuff to say the least.

I am starting to realize the existing condition may be "livable" providing we can make modifications (adding the bottom bars for additional flexural stresses at web yielding and preventing crack propagation)and as BA indicated make the safety factor against collapse improved dramatically.

The question I still have is can the welding take place to add the bottom cope bars in the as is stressed condition safely?

Will the high tempertures during welding reduce the existing web capacities to a point below the applied loads. I have looked at Blodeggett's (sp) welding text (great reference) and it indicates that the temp effects on strength are short lived, over a short time period and do not extend very far. The comments may have been made on fillet welds and not PP welds, I need to revisit the text tomorrow.

I still have more mulling over to do and to consider if the double plate BA was thinking about would be a better approach for welding in the as is condition.

Thanks for all the input.

 

[paddingtongreen]

If I remember correctly, they have to remove and replace the refractory lining periodically, that should provide a temporary reduction in load. Again, from memory, on the big vessels, it takes several weeks to strip out the worn brick and lay the new lining.

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[ToadJones]

BA-
yield would be a better word....
Having said that, if the thing is already overstress, at might be already yielded.
I am NOT trying to argue here, believe me.
But, if you were standing straight up and down and I tied a rubber band to the back of your neck and the belt loop of your trousers, it wouldn't stretch until you bent over to pick up the morning paper.
We have done a lot of industrial demo jobs using pancake jacks and shoring towers in situations similar to this. As mentioned above, the jacking system can pretty accurately gauge the load being lifted.

You may also consider welding diagonal stiffeners on the web as a stirrup.

 

[BAretired]

ToadJones,

I much prefer to shore the load during welding, but this is not an option in the present case. If the coped web is 30% overstressed, it has not reached yield strength, but it is close to it.

If welding is to be performed while the beam is under load, it is preferable to (a) reduce the load as much as possible and (b) locate welds in areas of least stress.

The risk of failure during welding must be carefully assessed.

BA