PXC
Structural
- Aug 21, 2003
- 40
I would appreciate some feedback on a method I have used to design some reinforcement for a complex shaped reinforced concrete structure. (It is 9x9 m plan, and has a 7 m diameter cylindrical hole through it with vertical axis. It sits on rock excavated in a sperical shape so that the depth varies from 1 m in the corners to 3.5 m around the cylindrical hole). I have modelled it using “brick” elements. I have modelled the concrete as an isotropic material, and used a linear elastic analysis. It is subject to potentially very large vertically downwards forces, but the shape results in compressive stresses throughout most of the concrete. There are relatively low tensile stresses (2.2 MPa maximum) in some areas that I need to design reinforcement for.
I initially designed the reinforcement for a “strip” through the critical section as a short column that was subject to axial compression with bending about 2 axes, and it required 15+ bars (28 dia) in the tension area, plus shear reinforcement. I then analysed it as a deep-beam and used a strut and tie method, and this required only 1 bar and no shear reinforcement! This latter approach was so simple and seemed more appropriate so that I proceeded to do use it for other sections.
Basically, what I did was to take a slice through the concrete at a section and integrate the normal tensile stresses over that slice to give a normal tensile force. Then I simply designed the reinforcement for that force.
My difficulty with this approach is reconciling the FEA results with real concrete behaviour. The tensile strain in the concrete in the model is very small so that in practice, the concrete must crack for the reinforcement to develop its full capacity. I can accept the validity of this approach for a simple beam, but it is harder to understand for a complex shaped structure. In this case, I assume such cracks could result in significant redistribution of forces. This is possibly no different to yield line theory for slabs for instance.
I would appreciate some feedback on the validity of my approach. And, if it is valid, then it raises the question: Why we don’t use this approach for the design of a (short) column?
I initially designed the reinforcement for a “strip” through the critical section as a short column that was subject to axial compression with bending about 2 axes, and it required 15+ bars (28 dia) in the tension area, plus shear reinforcement. I then analysed it as a deep-beam and used a strut and tie method, and this required only 1 bar and no shear reinforcement! This latter approach was so simple and seemed more appropriate so that I proceeded to do use it for other sections.
Basically, what I did was to take a slice through the concrete at a section and integrate the normal tensile stresses over that slice to give a normal tensile force. Then I simply designed the reinforcement for that force.
My difficulty with this approach is reconciling the FEA results with real concrete behaviour. The tensile strain in the concrete in the model is very small so that in practice, the concrete must crack for the reinforcement to develop its full capacity. I can accept the validity of this approach for a simple beam, but it is harder to understand for a complex shaped structure. In this case, I assume such cracks could result in significant redistribution of forces. This is possibly no different to yield line theory for slabs for instance.
I would appreciate some feedback on the validity of my approach. And, if it is valid, then it raises the question: Why we don’t use this approach for the design of a (short) column?