Seismic analysis steel structure on concrete columns 1

[DomDunk]

Dears,

I am designing a steel industrial warehouse with span of 40 m. The steel structure is made of truss beams and columns supported on concrete columns 5 m high.

I understand what is the seismic behavior, but I can not find the design method in ASCE 7-16, especially with respect to the seismic dissipation system (R).

How do you calculate these structures?

Thanks,

Daniel

 

[structSU10]

To clarify, do you have a trussed moment frame of some height sitting on 5m high concrete columns? no intermediate floor level? If that is the case, it might be a steel moment frame sitting on cantilevered concrete columns, which doesn't sound great.

 

[DomDunk]

It's like you write it. 5 m of columns in cantilevered concrete, 7 m of truss columns, and the truss beam, without intermediate levels. The client needs the columns in particular because there are machines that can hit the columns, and he does not want use more concrete.

The modal analysis of the steel structure supported by the concrete columns and the model with only the steel structure shows that the vibration modes do not change. The concrete columns have a very large rigidity with respect to the rigidity of the steel structure.

I am designing the steel structure under the design earthquake (it was also difficult to find a seismic system for a truss structure with this span and in risk category D). I calculated the base plate with the overstrength seismic forces. I'm thinking in a design calcualtion of concrete elements based in the overstrength seismic forces.

Are there parts of the standard that help with design criteria in this case?

 

[BridgeSmith]

Due to the height of the columns, and that they are cantilevers, I assume plastic hinging of the columns will not be allowed. I suggest considering a composite system with a steel column all the way to the foundation, with a concrete column formed around it. For the seismic loading, you would ignore the concrete, a just design it as a steel frame, assuming the columns would have to be repaired after the design seismic event.

 

[DomDunk]

I like the mental exercise you propose. I think that's the way!

But, in the case you propose, the dissipation occurs in the most loaded section of the steel columns ... The concrete modifies the location of the seismic dissipation in the column, because the lower part, where there is concrete, is much more rigid and resistant ...

 

[structSU10]

What is your seismic design category? Is this high seismic, or are you able to use 'steel system no specifically detailed...'? If low seismic, extending steel all the down, and doing a equivalent lateral force analysis, with any composite action you wish to count on, seems the way to go. If high seismic its a different issue.

 

[DomDunk]

Seismic Design Category: D

Steel ordinary moment frames: With permitted height increase. R=2.5

I have designed, based on the resistance of the elements, the connections in the connection zone between column and beam.

For those spans, with truss structures, I can not find a system that suits better.

 

[structSU10]

I would think designing it as a full height steel frame, bringing the steel to the foundation, and then treating the concrete as sacrificial, may be easier to justify. The mixed system, with a change in material, is difficult to know how it will truly react - you may also need to consider it as a different defined SFRS as currently set up, as the concrete columns below screw up the detailing / force dissipation philosophy that the R value in ASCE 7 is based on.

 

[BridgeSmith]

"...the lower part, where there is concrete, is much more rigid and resistant."

Yes, the concrete columns will make that portion stiffer and stronger, which will draw more of the moment, but it should be somewhat conservative to design the steel frame ignoring the strength contribution of the concrete, assuming it will crack and break away from the steel column during the seismic event. There will be some residual strength and stiffness from the concrete, but the detailing of the reinforcing and analysis of the stability of the structure for the 2nd mode shape and beyond, becomes more complex if you want to count on it for strength or stiffness.

 

[DomDunk]

I agree with both of you.

Anyway, the criteria for my design were the following:

- ensure that the columns in particular remain in elastic field (overstreght seismic forces)
- guarantee that the columns in concrete have over resistance with respect to the steel structure
- ensure that in the modal analysis the steel structure has more movement with respect to the concrete structure to concentrate the dissipation in the steel
- Evaluate that a modal analysis of the steel structure supported by erroje links "equal" to those of the complete model. In this case, the rigidity of the concrete columns would not greatly affect the seismic behavior of the steel part

The truth is that I would need a push over (I think it is the future of the seismic analysis to define R).

In the next project I will take your comments into account.

Thank you,

Daniel

 

[JAE]

OR.....

Take HotRod10's idea but separate the concrete from the steel columns.

The concrete's sole purpose in life is to protect the steel columns - it doesn't have to be structural at all but rather a surrounding shield protecting the column inside.
Sort of like a large wrap-around bollard.

So provide a soft wrap, or other means of separation and create a sort of concrete donut around the steel column.
Allow enough gap between them to accommodate your seismic deflections at the top of the concrete shield.



Check out Eng-Tips Forum's Policies here:
faq731-376

 

[DomDunk]

We came to the same conclusion with my colleagues in the calculation department.

Thanks

 

[KootK]

I trust that the use of braced frames, such as chevron etc, has been precluded for some reason? 13m is a tall moment frame column in any material.

 

[DomDunk]

In the longitudinal direction I used braced frame. In the transversal direction (40 m span) I can't use braces for the geometrical concept of the warehouse...

 

[BridgeSmith]

If you isolate the inner steel column from outside 'bollard', you have many more options for the outside - timber, fiberglass, etc. that will be more forgiving, and possibly more durable than concrete.

 

[WARose]

Seems to me like it would be a 2 parter for the seismic load: do the steel, then the concrete. (With the concrete part done as a cantilever special RC frame, which is allowed up to 35 feet in height in SDC D.)