Stiffness modifiers in buildings where steel and reinforced concrete are combined. 3

[austinb98]

How do you think I can adjust the stiffness modifiers when designing with linear analysis in hybrid structures where steel and reinforced concrete are used together?

First of all, stiffness modifiers given according to certain earthquake ground movements in TBDY-2018 are used for reinforced concrete elements. Here I want to use the general analysis method for steel elements. I am asked to take the axial, bending and shear stiffness of all elements as 0.8. The purpose of this is to include p-delta effects in linner analysis as an estimate of non-linear deformations. So, in this case, can you answer my questions below?

*While the stiffness modifiers we use when designing the steel elements in the hybrid high-rise building are set to 0.8, do you think it is necessary to multiply the values ​​given for the reinforced concrete elements by 0.8? I think there is no need for such a thing. Because we use stiffness modifiers for reinforced concrete elements, which are required by the regulations in case of an earthquake.

*Well, steel members will lose rigidity during an earthquake. In this case, if I use stiffness modifiers only on reinforced concrete elements, I have to design the steel elements for much larger loads than they actually are. However, there is no stiffness modifier recommendation for steel elements in our regulation (TBDY-2018). When I operate with AISC 360-10, I change the stiffnesses to 0.8 with the general analysis method. However, this is incorrect. Because the values ​​here are given only for p-delta effects. Are there any stiffness modifiers that I can use on steel elements such as reinforced concrete elements during an earthquake?

For example, in our regulation, the stiffness modifiers of walls for DD-2 earthquake motion are given as follows;
F11=0.5, F22=0.5 F12=0.5 M11=0.25 M22=0.25
In this case, in a composite column (concrete filled steel tube) in the same building
What stiffness modifiers should I use?

 

[JoshPlumSE]

This is an inherent flaw (in my opinion) of a very material centric analysis requirement (like AISC's Direct Analysis Method). That specification doesn't have ANY applicability to other materials, only steel.

I'd argue that the 0.8 stiffness reduction (from AISC) is supposed to apply to your concrete elements as well.... at least for the load combinations which you are using for steel design. The argument here is in the User Note in section C2.3 of the code. If you do NOT apply this reduction to other non-steel stiffnesses, then your steel forces and moments will be under-represented.

That being said, I would also argue that the load combinations that you use for concrete design do NOT need that additional 0.8 stiffness reduction. I'd have to go through the concrete code to decide whether I apply the 0.8 reduction to steel members for the load combinations that are used for concrete design. That would certainly be conservative (because the concrete would be relatively stiffer and attract more load).

 

[centondollar]

Steel doesn't typically crack or lose large amounts of stiffness from formation of plastic hinges at joints. Concrete cracks and loses large amounts of stiffness (particularly beams, less so shear walls and columns). Therefore, applying more or less aribitrary stiffness reductions to steel members doesn't make any sense to me, since modelling steel plasticity is not esoteric or riddled with uncertainties.

Stiffness modifiers are the wrong way to go about this if you want to do performance-based design. For such analysis and design, I would suggest non-linear time-history analysis using e.g., non-linear shells for shear walls and non-linear beams (define a normal-force dependent moment-curvature relation) for RC beams, columns and steel structures.

 

[austinb98]

Stiffness modifiers are the wrong way to go about this if you want to do performance-based design. For such analysis and design, I would suggest non-linear time-history analysis using e.g., non-linear shells for shear walls and non-linear beams (define a normal-force dependent moment-curvature relation) for RC beams, columns and steel structures.

I will do nonlinear analysis for performance evaluation. However, I design with linear analysis for dimensioning the elements. (Required by the regulation) My question is about the stiffness modifiers we use in linear analysis. As you know, in this analysis, we assume that the reinforced concrete elements crack. How accurate do you think it would be to think that the steel elements will never lose their stiffness? Because in this case, we increase the loads on the steel elements compared to the situation where we reduce their stiffness. Do steel elements not lose any stiffness during an earthquake?

 

[Just Some Nerd]

Do steel elements not lose any stiffness during an earthquake?

Through what mechanism would a steel member lose stiffness?

----------------------------------------------------------------------

Why yes, I do in fact have no idea what I'm talking about

 

[Greenalleycat]

Interesting one. I have no familiarity with your standards so trying to draw parallel to ours
It looks like your modifier of 0.8 is an inverse of the way it is dealt with in our codes
We have deflection modifiers that we apply to our calculated deflections to represent the simplifications in the analysis methods
So instead of stiffness * 0.8, we use deflection x1.2 etc
Perhaps thinking about it like this would help answer the question - in this case I would apply the 0.8 factor to everything as it is a fudge on ANALYTICAL inaccuracies, not material (if I understand it correctly)

We also still apply material specific reductions e.g. cracked section modifiers on concrete
These should still be applied in your case, IMO, appropriate for the level of load on each element etc in addition to the 0.8 factor

Disclaimer again: I have no familiarity with the standard so I could have misinterpreted something

 

[austinb98]

Interesting one. I have no familiarity with your standards so trying to draw parallel to ours
It looks like your modifier of 0.8 is an inverse of the way it is dealt with in our codes
We have deflection modifiers that we apply to our calculated deflections to represent the simplifications in the analysis methods
So instead of stiffness * 0.8, we use deflection x1.2 etc
Perhaps thinking about it like this would help answer the question - in this case I would apply the 0.8 factor to everything as it is a fudge on ANALYTICAL inaccuracies, not material (if I understand it correctly)

We also still apply material specific reductions e.g. cracked section modifiers on concrete
These should still be applied in your case, IMO, appropriate for the level of load on each element etc in addition to the 0.8 factor

Disclaimer again: I have no familiarity with the standard so I could have misinterpreted something

Thank you for your answer. Yes, in our regulation, we use 0.8 like you to increase the displacements in steel elements. (Control and application of P-delta effects)

However, the problem was actually this, in hybrid structures (steel + reinforced concrete), there is a decrease in the stiffness of reinforced concrete elements (cracking) during an earthquake, and we use stiffness modifiers like you. (in linear analysis) However, in this scenario, will there be a loss of stiffness in steel elements? That was basically what I wanted to ask.

As far as I understand, when applying load combinations for the design of steel elements, I should reduce the stiffness of reinforced concrete elements as follows,
Reinforced concrete frame column:
F11=0.7*0.8


For steel elements,
Steel column;
F11=0.8
In this way, while not disrupting the load flow, I will have made the desired increases for the control and application of p-delta effects in steel elements.

For the load combinations I will make for reinforced concrete elements;
Reinforced concrete frame column:
F11=0.7


For steel elements,
Steel column:
F11=1
It would be correct to use it as. At least that's what I understand from the comments.

 

[Just Some Nerd]

If my understanding of Greenalleycat's post is correct, I'd probably opt to apply ordinary modifiers to concrete with no modifiers to the steel, and multiply the output displacements. The inverse of a 0.8 stiffness modifier would be x1.25 I think, if I'm doing my math correctly.

----------------------------------------------------------------------

Why yes, I do in fact have no idea what I'm talking about

 

[austinb98]

To summarize the situation and simplify the question, let's first consider a building made of steel columns and beams and reinforced concrete walls.

*First of all, when designing reinforced concrete walls, we apply the stiffness reductions given in the regulation. We do not make any stiffness reductions for steel elements. (Linear analysis) We also consider whether p-delta effects will be applied to the structure according to the obtained data.

*Secondly, when designing steel elements, if the entire structure was steel, we would change the stiffnesses of all elements by 0.8 and increase the deflections in the structure. (In this way, the load flow would not be disrupted) However, since the structure is reinforced concrete + steel, we do not reduce the stiffnesses of steel elements. We use the stiffness modifiers given for reinforced concrete. We apply deflection modifiers to the obtained deflection data. In this way, we increase the displacements by preserving the load flow.

If I am wrong here, please correct me*

However, in my construction, there are columns with concrete filled steel tubes inside. The concrete in these columns will crack and the rigidity of the columns will change. In this case, it would not be a correct approach not to reduce the rigidity of these columns.