Seismosoft | Earthquake Engineering Solutions

Hello,

Use of ccs "composite circular section" is brought up in the forum for the concrete filled steel tube or steel jacketing application. I have some concerns and do highly appreciate the help for better and accurate understanding.

1) What is the interface between the steel tube and the concrete core inside for the "ccs", if there's any in SStruct? Is the use of link element mandatory for this case to capture the relative slip between the tube and the concrete inside?

2) Is the constitutive behavioral model for an identical column one with transverse stirrups and one encased in the steel tube similar? Does the SStruct include the confining effect from the steel tube explicitly, or needs to be introduced via the confined concrete model seperately? Where the latter seems right to me!

3) Is implementation of SStruct appropriate for a case of Concrete filled Steel Tube (ccs) wrapped with an external layer of FRP tube (chs-Circular hollow section)in the plastic hinge? In other words, is the composite behavior between steel tube and FRP and their different corresponding dilation properties incorporated or not?

Trying to simulate a Reinforced Confined Concrete Filled Steel Tube (steel tube filled with concrete and wrapped with FRP outside) brought up the above questions. I am doubt-full about the method that I need to utilize the confining models and simulating the local behavior. Thanks for your value able technical comments in advance.

Hi fakharifar. I'll relay my understanding of this element, which I have used.

1. I believe that there is no slip between the tube and concrete at the interface in the model.

2. I believe that the concrete behavior would be identical whether the confining is provided by spirals or a tube. The section behaves differently however due to the tube. I believe that you need to model the concrete inside the tube with appropriate con_ma with the equivalent confining factor and other parameters. I do not believe that confining effects are implicit in the ccs section. This is the way I have used it.

3. I suppose that you could use some sort of transformed properties of an equivalent tube to mimic the combined effects of steel tube and frp wrapping. I am imagining an artificial material for the tubing with properties calculated using the modular ratio (Efrp/Esteel) to transform.

Press on fakharifar.

Thanks dear huffte for the comment. I have been skeptical a little as cited below. Sorry for the long post in advance.

1) I agree with you that there is no slip between the concrete and steel tube in the "ccs" section. I did a simple trial test, a composite circular section with a thick wall under axial and lateral load simultaneously and output the strains at different locations (concrete core and steel tube). The strains are equal for the core and on the steel wall. Then, what if I use the regular reinforced concrete circular section and locate it at the center of a new circular hollow section in the elements connectivity.

This would raise in two concerns:
1-a) What is the interface between these two separate cross sections that are modeled as a composite cross section (hollow steel tube outside and the reinforced concrete inside)? What are the main differences between this case compared to the case that the "composite circular section" already available in the SStrcut is used?
The help manual wouldn't give me in depth answer for this. The sensitivity analysis between these two cases didn't make conclusive correlations to draw some modelling hints for this case.

1-b) When the method of using two separate cross sections that are modeled as a composite cross section is used, which calibrated link element could be used for the slip? I could ignore the slip, however it'll rise the error since part of energy dissipation by the slip is missing.

2- I have used the same method calculating the right confining factor (con_ma)provided by the steel tube and assigned it to the concrete inside the tube. Then, I can't justify how the tube is affecting the results. In other words, why should the "ccs" section be used if it is not affecting the overall behavior of the encased concrete inside it, and the confinement factor is all that needs to be defined? There should be something considered by SStruct for this section different, I guess.

3- I strongly agree with the transformed properties idea for the FRP wrapped steel tube, however I was concerned for the nonlinear response of such modified material since it is constituent of two totally different materials.

How about using the same method as described in the "part 1". Just using the:

Regular reinforced cross section located at the center of two hollow tubes (Steel tube + FRP tube) accordingly in the element connectivity tab. In this way the two constitutive material behavior for Steel and FRP are modeled. However, post processing of a column made from this system gave inconsistent node displacements at the same level for the three nodes belonging to concrete, steel and FRP tube.

Thanks again for the response for better insight.

Regarding how the tube affects the results for ccs with confined concrete versus a con_ma core by itself - given that the confinement of the tube is not explicitly accounted for in the algorithm, you still have the fibers and their associated stress-strain properties adding to the section strength and stiffness. So strain compatibility is maintained at the interface, as mentioned. If two separate coincident elements - one a tube and the second a con_ma core - then no such strain compatibility is imposed.

Thank you much Huffte.

Any help/idea on the relative slip modelling between the steel tube and concrete infill is highly appreciated.

Much thanks in advance.