Instability of very deep cast-insitu post-tensioned (PT) concrete continuous unrestrained transfer girders

[user277418]

Hi

A case:
A transfer structure of a building. Three parallel post-tensioned (PT) continuous transfer girders are holding several storeys frome above. The girders are formed from a web WIDTHxHEIGHT=0.8x2.5m + a PT slab 0.28 m thick on top (i.e. a composite T-section. Going to be casted in 2 stages). The girders are sitting on top of relatively narrow and slender columns 0.58x0.58 m. Step of the columns is about 10 m along the girders. Every girder has 4 spans. Since the girders have so deep webs (almost 1 storey height) I am considering provision of transverse diaphragms of the same height as the girders over the columns (transversely to longitudinal axis of the girders). By analogy to the diaphragms over piers in construction of bridges. I am worried about out-of-plane instability of the girders since they are exceptionally deep, continuous and have unbraced bottom of the web in the zone of hogging bending moments.

A problem:
For me the provision of the diaphragms seems logical. However I have a very strong suspicion that my project manager would not appreciate it unless I am going to bring him some solid prove that they must be provided (HVAC engineer has already reserved the space for passing his ducts). Eurocode is almost silent regarding instability of reinforced concrete beams. LTB of steel-concrete composite girders restrained by cast-insitu concrete slabs may be of some help, but is not directly applicable. Most of available publications are discussing instability of precast prestressed girders during construction or lifting. The publications can hardly be applied to my case since my girders are going to be fully supported during construction. My main concern is about final state.
As an alternative to the diaphragms I am thinking about an option of designing the column to girder joints for a notional lateral force (NLF) applied normal to the girders plane. The NLF is going to be equal to a maximum hogging bending moment over the columns divided onto depth of the girder. Additionally to the amplified bending moment according to EC.

Answers I am looking for:
- Do you consider such case to be a problem at all? Regarding instability of RC girders
- Do you know any publications describing design of similar cases or behavior of continuous reinforced concrete beams of such depth? Seems to me, design of cast-insitu bridge girders may have some similarities.
- What do you think about the notional lateral force instead of the diaphragm? I prefer diaphragms

Any opinion is appreciated,
Best regards

 

[HTURKAK]

Step of the columns is about 10 m along the girders. Every girder has 4 spans.

What is step of the column? Can you provide a sketch?

I am worried about out-of-plane instability of the girders since they are exceptionally deep, continuous

You may provide transversal ( also AKA cross beams ) as necessary.

I have seen some transfer girders made up of steel plate girders . Consider steel girder option.

 

[user277418]

What is step of the column? Can you provide a sketch?

The step of the columns is distances between the columns c/c along the girders. By other words spans/bays of the girders.
And the steps are actually about 15 m instead of 10 m. My mistake.
Will add the sketch a bit later.

You may provide transversal ( also AKA cross beams ) as necessary.

The diaphragms, I were mentioning, are the cross beams basically.

I have seen some transfer girders made up of steel plate girders . Consider steel girder option.

Because of environment and vibration requirements steel is out of discussion. At least for now.

 

[KootK]

Do you consider such case to be a problem at all? Regarding instability of RC girders

Nope. I agree with your manager that this is unnecessary so long as:

1) The tops of the girders are restrained laterally by a floor plate and;

2) The bottoms of the girders are restrained laterally by a floor plate.

Normally, I would expect the slabs to be themselves laterally restrained by whatever is serving as the global, lateral bracing system for the building (shear walls, moment frames etc).