Steel liner to concrete structure - thermal loads

[bkal]

Hello,

I have been struggling with this for some time and I can only conclude that some of my basic assumptions are wrong.

We need to provide a steel liner (3-5mm thick steel plates) to a single room in a large reinforced concrete structure (~1m thick concrete). The steel liner plates will be welded onto flats / steel profiles embedded (anchored) into the concrete on a grid approximately 1m x 1m. The liner goes only half way up the wall, and then stops, it is covering all the floor area of the room.

There is a possibility that the temperature in the room can suddenly increase by 100deg C, which means that the temperature of the liner is going to be (for some period opf time) significanlty higher than the temperature of concrete it is connected to. Concrete is also restrained from free expansion by surrounding concrete (only that single room in a large building is lined and can experience high temperature). The issue is that the in-plane force the liner puts into the embedded elements is very high, which is causing problems for its anchorage.

While it can be argued that the embedded elements do not experience any high net (in plane) force if it is supporting two liner plates, the top embedded elements in the walls have only one plate attached to it so will see a high net force.

I would appreciate if anybody can advise if I am getting my physics / mechanics wrong, or if there are any clever solutions to this issue.

Thanks in advance.

 

[ishvaaag]

You "just" need to analyze the tthermomecanical evolution of the model composed with every steel and concrete placed therein. Transient evolution should also be considered since it will likely be worse than the stabilized steady state.

Quite likely Algor and most of the multiphysics packages can deal with this. Such way you may get a more refined and likely somewhat lower statement of the averaged stresses you refer to than in just a simplified elongation approach, what can be useful to ascertain if the anchorage provided is enough. <

This kind of deformation can be seen in frying pans, that buckle, something you are likely to see if only 3 mm thick on 1000x1000 mm restraint grid. You may independize to some extent the liner and anchorage from the concrete backup to just ease stresses. Other thing is that this be acceptable.

 

[hokie66]

I imagine the liner plates will buckle before the anchors fail. I would look at allowing the plates to expand freely. Perhaps weld them only on one side, with the other sides retained in a slot?

 

[bkal]

Thanks for input. We now have results showing that the plates will indeed be hotter and would experience higher tenedency to expand comapring to the consrete they are connected to. Sliding joint is probably not an option as the plates are tiled (as they cover quite a large area), so there might be a problem of ensuring leak tightness at the sliding joint.

We have also considered claiming buckling as a relief mechanism and it works fine as long as the plates tend to buckle away from the concrete - however the problem is if they tend to buckle into the concrete, then their buckling stress becomes more significant.

Again, thanks for your thoughts.

 

[ishvaaag]

Quite likely, only pre-stamped plates with some bulge outwards from the concrete face may warrant the relief. It then becomes necessary to check how the bulge stands any hydrostatic pressure present, as treading on the plates. In this idea, you prefabricate pre-stamped plates and weld against the backup grill anchor; maybe even studs at nodes can be enough. Equipment (if there is some) needs then be supported at nodes.

We can also start searching the web for the matter.

http://www.iasmirt.org/iasmirt-2/SMiRT19/S19_FinalPapers/H01_2.pdf

I am leaving now, but will search more on the matter later.

 

[SAIL3]

In my own experience, it is not a good idea to try and resist thermal expansion loads if it can be possibly avoided..
I would tend to let the liner float....
so the plate may buckle..so what?
it would not be a structural problem..at best, it would be a cosmetic problem..at worst, it could be a leakage problem..
to install I would use backing plates on the plate section you indicated and use butt-welds to join the plate sections.
the backing bars placed against the conc wall will provide a gap of 3-5mm for the main plate sections.
where the plate ends half-way up the wall I would provide a stiffening lip on the plate and add a flexible seal, if leakage is a concern....
if you are concerned about the added stiffeness of the corners and the thermal loads they might produce then you could add a nominal radius to the plate in those areas .
from your post I am not sure what the function of the liner actually is...leakage prevention??...the conc walls themselves may
handle that...corrosive liguid???..well if you ever do develop a leak in this kind of set-up, it would be very difficult to locate it!!

 

[graybeach]

I might try a thicker plate with a closely spaced (150 mm?) grid of shear studs welded to the embedded side. I'm envisioning the inside face of the concrete being subject to tension so I'd place reinforcing for that. The outside face of the concrete would be in compression.

But what about walls above the steel faced concrete? maybe some sort of expansion joint?

 

[ishvaaag]

If you have controlled the problem of tensile stresses in the concrete through reinforcement, you should be able to do the same beyond the liner. I mean, the reinforcement is already restraining the unwanted cracks where the cracking forces are at the biggest value, so from there only lower strains should be expected.

Respect my idea, really it would need a cross shaped bulge, free to go upwards in the bulged parts, then seating on the concave (from the interior) parts. This may not be entirely suitable to proper prevention of leakage through welds.

I have found a number of articles mostly in the nuclear industry related to liners but none as good as to be quoted for the question. Will try more later, I hope.

 

[TXStructural]

Steel liner? Have you considered a polymer lining instead? It's a dumb question, but I hope it has been asked before. Also, would you consider fewer anchorages so that the thermal expansion is not restrained, but allowed to occur? I would expect at least 1/16" per 4' width at the 100C temp change.. I would think that this amount of oil-canning across a single sheet would be acceptable.

 

[Once20036]

I can't think of any rooms that would be subject to "sudden increases" of up to 100 degrees C, but assume that if it's a sudden increase it is also associated with a sudden decrease.

Depending on the time lapse between the sudden increase and sudden decrease, there may not be time for the steel to "absorb" 100 degrees worth of energy and expand to stress the concrete.

If the heat is only in place for a minute or two, I would expect the surface of the steel to heat up, but not nearly as much as you move toward the center of the concrete.

 

[ishvaaag]

The closely spaced studs look a promising idea in that
1. studs may restrain the tributary unbalanced actual forces
2. once effectively made composite, you may understand the sections as under self-equilibrating thermal stresses; that will give you shears at the interface required to be taken by the studs and then the stresses in the concrete to be counteracted with reinforcement; or as a FEM analysis would give.

In short, you have devised a design where all the required parameters seem to be reasonably derivable, and so it is a go ahead. It is unlikely that 100 deg of temperature may cause strains of such level that no proper control of the behavior be at hand; the shears at the interfaces of composite beams should be a worse example of what can happen with the kind of the solicitations at the interface, and they still work.

 

[ishvaaag]

Just an example of one attempted simplification of the steel liner with studs in a mat, by (Mathcad) hand.

 

[ishvaaag]

I don't see the file, I try again

 

[ishvaaag]

By whatever the reason the file does not upload. I reboot and try again later.

 

[ishvaaag]

Again

 

[ishvaaag]

Hmm, it seems impossible to upload this small (1 page) pdf file. Please review if something is wrong with the storage service and will try again.

 

[ishvaaag]

A new try

 

[ishvaaag]

Again

 

[ishvaaag]

Let's try Ron's "second try principle"...

 

[ishvaaag]

Not working for me