Strut-and-tie modelling of steel structures?

[bugbus]

This is more of a hypothetical discussion.

In theory, is there any reason why the principles of STM could not be applied to the design of structural steel (i.e. no concrete)?

I am envisaging the design of nonflexural steel members, brackets, corbels, deep beams, connections, corners of frames, and so forth.

I suppose one of the main differences would be the need to check for buckling of plate elements in compression, but other than that I imagine the remaining checks might be pretty similar to what is normally done for RC.

Another point would be that FEA is quite suited for this sort of thing since the steel can be well modelled by plate elements with pretty well defined material properties. Concrete obviously has the added complexities of cracking, discrete reinforcement bars, more brittle behaviour in compression etc., which is where STM comes in handy.

 

[KootK]

I would say that yes, similar principles can be used in steel design with the right attention to steel's unique nature (buckling as you mentioned). I've done this myself for the design of stiffened seat brackets. I'd also consider the tension field method of steel beam shear design to be an example of this.

 

[rowingengineer]

I would say yes as well, however I am more of a fan of FEA especially if you have bolts in tension.

With the software programs you can get today that help with connections, brackets, and corbels with full finite element programs tracking the plate bending and bolt forces is easy.

 

[KootK]

Which FEA programs are your favorites for that rowingengineer?

 

[rowingengineer]

For connections I have to say idea-statica

https://www.ideastatica.com/

has my attention at the moment. If I was to put together a FEA program for connections this is how I would do it.

Other than that I use a Australian program Strand 7 but it is just a general FEA environment, I am sure there are many that are similar available.

Please note that I do not get any royalty or similar for my love for these programs.

 

[MegaStructures]

+1 for ideastatica! An exciting program for sure.

KootK beat me to it I think, but I came to say that there are two main steel "components" that come to mind that warrant the use of the S&T method for steel.

1) Stiffened Seats

In my mind these are analogous to concrete corbels. The compression field travels through only a small portion of the plate from the point of load application diagonally down to the support, and the tension field travels through the top of the plate back to the support. Without pulling the book out I believe Steel Structures Design and Behavior by Salmon recommends this method.

2) Web Shear Stiffners

Again, analogous to deep beams in reinforced concrete with a T/C force travelling diagonally in between stiffeners.

Edit: I also don't see the buckling tendency of steel to be a limitation to using the S&T method. The method simply is used to determine where compression is flowing in a component. At that point that section of the steel would be designed as a small column. Though I do wonder what the effective length factor would be for a "column" supported on all edges, surely that is listed in the Salmon reference above, or Roarks.

“Any idiot can build a bridge that stands, but it takes an engineer to build a bridge that barely stands.”