[STM] unsymmetrical design forces in strut and tie.

[TewitC]

Hi I'm currently self-studying the strut and tie method. I get the general concepts and they all make sense. But then I started noticing how sensitive the validity of the strut and tie model is. For example:

In a very simple strut and tie example I have seen all over the internet (see attached image), can be applied to deep beams and pile caps, the applied load is divided into "A" and "B". The struts on left and right side will have the same amount of compression force because of the symmetry. HOWEVER, if "A" and "B" are not equal, even the slightest difference, the model loses symmetry and becomes unstable. The tie force required when calculated at the two nodes would be different. This cannot be as the tie force is constant in truss analysis. Therefore, to solve this, another node must be added along the tie to compensate that difference.

While I do believe this is the right approach, I rarely see an example that takes this into account, especially when it comes to a more advanced application such as in 3 or 4-pile pilecap. In those two in particular, I constantly see statically unstable STM models that rely heavily on the symmetry of the applied force, something that can't be achieved in practice. This is even worse when it comes to pile caps supporting bridges, which surely will have some bending moment transferred from a column, or even govern the design in some cases. Also, the design results can be very critical, ignoring the moment results in some STM members missing, which eventually leads to improper details in those areas (i.e. insufficient column starter bar anchorage or missing reo. in certain directions)

So basically the question is: Do I miss something? Is there some kinds of simplification or rule of thumbs involved? As I said I'm just starting to self-studying this, so my understanding might be off, please correct me.

 

[Tomfh]

This was discussed a few months ago:

https://www.eng-tips.com/viewthread.cfm?qid=482826

 

[TewitC]

Tomfh

Thanks! I have read through that post. The comment by KootK on 7 May 21 was what I was thinking about, good to know that someone agrees. But I still wonder what designers do in practice when have to deal with this asymmetry, especially in a more complicated cases. Because in that post, it seems like the procedure is not something so universal that every engineer agrees on. Also, I have asked around my office and seems like no one really dealt with or thought about that before, strange but true story (a small firm though).

 

[KootK]

But I still wonder what designers do in practice when have to deal with this asymmetry, especially in a more complicated cases.

In a lot of cases where a potential asymmetry would cause problems, I suspect that condition is often being handled poorly by designers. It's been handled poorly by me in the past.

In theory, solving this issue is no more complex than simply checking one's design for all of the applicable load cases. In practice, given that most STM is done by hand or with only basic, spreadsheet assistance, executing that becomes a tedious nightmare. Consequently, designers try to pare down the load cases to just the one or two that govern". That process usually serves us well but, for these asymmetry conditions, there seem to be holes in the procedure.

I've been noodling on what a reasonable, modified procedure might be. I'm currently thinking of something like this:

1) Using software, model all of your load cases. This requires the designer to actively tend to not just code load cases but, effectively, pattern loading.

2) Generate an envelope shear diagram and start with the assumption that conventional stirrups will be provided to deal with that shear at all locations, even when that seems unreasonable.

3) When removing stirrups from a zone within a beam and replacing that shear capacity with a concrete, go back into the model and identify the particular load case that produced that peak shear. And ensure that situation is dealt with properly in the STM modelling.

To me, the whole thing feels very much like how patterned, partial loadings on beams can produce higher shears at particular locations than the full loading condition does. It's that phenomenon that informs the approach that I've proposed above. If somebody's got a better one, I'm all ears.

 

[Brad805]

Good to see you back Kootk.

 

[KootK]

Thanks Brad805, it's great to be back. I've missed the gang and real life hasn't provided me with nearly enough intellectual stimulation.