Steel Beam Deflection with Same Equipment after Refurbishment

Metal fatigue?

That is the guess at the moment but if there is an in depth explanation of what caused this and where I can read more about it, it will be appreciatted.

Here's some links...

http://www.nps.gov/goga/historyculture/upload/Chapter%2010%20B.pdf

http://www.constructionknowledge.net/general_technical_knowledge/general_tech_basic_structural_design.php

Is the equipment vibratory?

To what previous measurement is the deflection being compared? or is it from a string line from end to end?

Was the deflection ever measured before it was re-painted?

What is the span of the outer beam?

If fatigue, there are usually some cracks, have you seen any?

I got the call today from the team at site - I will go there myself next week only. Everyone break now for Christmas. That is the problem - nobody ever measured it but at the span of 5.760 meter it never was according them so evident. There are only cooling towers on top.

20mm in 5.76 m is 1 in 288, which is a bit on the high side for design, but not unacceptable in itself. Does the deflection cause any problems? If not, I would estimate the steel stress, and if that is within limits, stop worrying about it.

As mentioned, fatigue could certainly be a factor. However, it may be possible that the refurbishing methods could have impacted the member strength. What method(s) were used to remove the rust? Was it mechanically removed? Chemically removed, if so what chemicals were used? And as mentioned, could it have been this deflected before the refurb?

Wire brushed and sanding paper - no chemicals as in the off season we seek work for the bottom line staff. 2 x layers anti rust - 2 x layers paint.

Thanks all for the reference - i am just curious but will have a look at it next week.

Was the deflection measured before the refurbishment, so that apples could be compared with apples?

Along with all the other good ideas put forth I would look at the re-fastening of the supported equipment. Was the same arrangement of spacers, washers etc used? Was it all torqued the same and was the fastening sequence the same? It is amazing what effect that little changes in assembly procedures can have in the final results of constructing something.

Please keep us posted of your findings.

The complete structure as "welded" back down in the 70"s. My at this moment stupid as it sounds, guess is that the load caused a "relieve of binding fibres" with off loading and then when it got loaded again it was being pushed just passed the elastic range. The beams are very old and we do not even have reference in any of our modern books anymore - it is 250 x 150 I-beams ( 11" x 6" ) Appearantly according those who could remember it was done by a local eng shop - the guy who did it is also dead.

Here in India permissble deflection for cantilever is L/350=5760/350= 16.4 mm.In your case it is 20mm which is not allowable. You can either support beam by Tie Bar or weld additional steel channel/plate on the top of the beam. Better consult your architect.

I suggest you perform a simple stress analysis to determine the beam stress when the structure was put into service and the safety margins. If you don't have access to the calculations, we may be able to help (unofficially) Fatigue could be part of your problem but I wouldn't rule out creep. Also, seismic or meteorological events that may have occurred could have increased the static load substantially.

I question whether the existing framing connections were re-used in the same manner after the beam was refurbished.

A beam with standard web clip angles, bolts, washers, and nuts ("simple support") end connections will deflect much more than a "fixed" end connection. Guarantee it under identical loading conditions....

Also, are you absolutely positive that any of the beams (having the same span lengths, or nearly so) did not get accidentally swapped? If this did happen, and their sectional properties are visually close (but not identical), the beam having a lesser Area, Sectional Modulus, and Moment of Inertia could have been inadvertently erection in the wrong location (not it's original location), thereby possibly causing a greater deflection under loading.

Just a thought.....I have seen it happen all too many times on a job site where beams were accidentally erected in the incorrect location; sometimes leading to catastrophic structural failure.

BTW, this should have been originally posted in the Civil Engineering Section, and not in the Mechanical Engineering Section.....

Fixed and pinned - solid weld connections.

I actually picked general section, why it came here I do not know.

The calc's can be done, thank you very much for the offer. I will go there next Wednesday for an inspection. It is a summer holiday break for us here.

Happy Christmas to all and thank you for the references and the comments.

the reason why is fairly straight foward. cross members on both sides of the units should fix the problem.

Before refurbishment little bend and after refurb big bend.

Machinery getting heavier after a refurb seems odd...

Were these water coolers in use before the refurb or were they idle?

If the water coolers were idle they may have been empty and now that they are back in service they are full of water and heavier.

They were operational before and after.

Did you flip the beam over?

Have you tried flipping it to allow any pre bend to be straightened out when loaded?

Do the legs pull in or push out when loaded? It has legs right?

Was some bracing omitted?

Wall - plse read my posts - it is solid welded as one side fixed other side pinned. the same machinery was removed and the same machinery was put back, the job it did before, it is still doiing. The bracings is the same and still in tact. It was wire brushes, sand papered and repainted by hand - no fynny chemicals was used etc, etc. The beams is old beams of the 1960's and was constructed at least 30 - 40 years back. It is standing outside in temperatures of 45 - (-3)°C with normal seasonal wind loads etc, etc. I think what should have been read before asking / making a sugeestion is now answered - and it is very bad practice to flip an opering beam to counter flex it a full load condition when it sagged out anyway.

I thought about it for the past few days and - I as I have not seen it yet anyway suspect that the steel was anyway loaded years back to it upper stress limit and the time over fatigue kicked in. I believe that it is by accident that it sagged now just 2 x years after repaint job.

I was not there yet and will be as said above there next Wednesday for a personal inspection as it is summer holidays here and it is tomorrow Christmas.

Sorry boss!

Just guessing sight unseen with anecdotal history.

You described the cleaning briefly. Is there any chance that it may have been cleaned more vigorously than you think (loads of scale) and the material is now significantly smaller? But then it was OK (reportedly) for 2 years...

It is OK - my mistake, I myself do not always read through all the posts. It was not very rusted - only surface rust and we did it to keep the lower employee group busy during the off season

I hope the site is in a lovely location.

Enjoy the field trip.

My observations :

- The structure must have been designed for static load. The cooling tower in fact is not absolutely a static load. It contains lot of water which is moving and also the top mounted fans which can add to the dynamic loads.

- You must check the original design of the structure. It seems just adequately designed to take care of the existing loads and that too in a New condition.

- Now the question is why did it did NOT BEND in the NEW CONDITION? I think, when the structure is new, it was just sufficiently strong to withstand the load of cooling tower. Now when the main structure becomes weaker due to ageing or fatigue,the structure associated with the cooling tower (the Load) in this case acts a load bearing member and does not permit the supporting members to deflect.

- I am very sure that if the deflection had been measured immediately after Removing the Cooling Tower, you would have observed the same amount of deflection what you are seeing now.

- Hope it explains and conveyed.

IMO, if this deflection issue is such a big concern then why not in the first place just replace or reinforce the steel beam(s) to acquire structural adequacy...that way you'll have a peace of mind!

CAPTMOOSIE

I was just asking out of curiosity - not expecting this kinda response - thank you for the brain leaders. In the end I will have it to be replaced - I am pretty sure about it but out of courtesy I am answering everyone as it is coming. Will do the inspection and then the calc's.

I think we can end this conversation now - thank you for all that was on it.

We may decide to annoy you with wild guesses in spite of the fact that you are content with what we have said , so far. One additional point. Are you certain that the deflection you observe is bending of the beam, or could it be that the beam's position is slightly shifted from its original location. During recent welding of the fixed end, weld stress may have shifted the angle of the beam from its original location so the pinned end is at a different elevation from original.

Not annoying to me but it looks like it is annoying others. Thank you -something I will keep in mind.

One thing that I failed to mention before was that you could retrofit the beam by installing diagonal struts (at 45 degrees) between, say at the 1/3 span points of the beam and the two supporting columns at the ends of the beam so as to reduce the span, and thus reduce the bending and shear stresses, and the positive (downwards) deflection.

Of course you'd have to install a secondary horizontal piece of steel between the columns where the new diagonal struts meet the column faces. This would be required to take-up and neutralize the compressive forces acting on the columns. Said horizontal strut is a secondary bracing piece and will be in tension.

This may be the cheapest route rather than replacing the original beams, and will help stiffen up the entire topside framing. All connections could be field welded, say with E70XX electrodes. All struts could be square tube steel due to their inherent higher Radius of Gyration, which will knock down the Design Slenderness Ratio (KLx/Rx, KLy/Ry) of the compression struts.

Just a thought that you may want to consider.....

Good luck, and please have a Very Merry Christmas wherever in the World you may be!

The thought so far is on that path - A new horisontal beam below that one with stubs upward, what must be considered though is that beam can/may/possibly is past the elastic point and on the plastic curvature, thus you must not actaually see the existing beam as a contributor to the strength of span support. Thus the new beam will have to be the "main" beam taking all the load. Thus a calc on the "new main beam" with all loads DL & LL - Tha is how I see it at the moment CAPTMOOSIE.

To all in the world - may there be piece in your hearts, may you drive safe and may God Bless you. A Merry Christmas and a Happy new year.